fix: simplify finish reason warning condition

.github/workflows/github-issue-tracker.yml

@@ -42,7 +42,7 @@ jobs:
 
       - name: Add pending label if in quiet hours
         if: steps.check_time.outputs.should_delay == 'true'
-        uses: actions/github-script@v8
+        uses: actions/github-script@v7
         with:
           script: |
             github.rest.issues.addLabels({

.oxlintrc.json

@@ -11,7 +11,6 @@
     "dist/**",
     "out/**",
     "local/**",
-    "tests/**",
     ".yarn/**",
     ".gitignore",
     "scripts/cloudflare-worker.js",

CLAUDE.md

@@ -12,15 +12,7 @@ This file provides guidance to AI coding assistants when working with code in th
 - **Always propose before executing**: Before making any changes, clearly explain your planned approach and wait for explicit user approval to ensure alignment and prevent unwanted modifications.
 - **Lint, test, and format before completion**: Coding tasks are only complete after running `yarn lint`, `yarn test`, and `yarn format` successfully.
 - **Write conventional commits**: Commit small, focused changes using Conventional Commit messages (e.g., `feat:`, `fix:`, `refactor:`, `docs:`).
-
-## Pull Request Workflow (CRITICAL)
-
-When creating a Pull Request, you MUST:
-
-1. **Read the PR template first**: Always read `.github/pull_request_template.md` before creating the PR
-2. **Follow ALL template sections**: Structure the `--body` parameter to include every section from the template
-3. **Never skip sections**: Include all sections even if marking them as N/A or "None"
-4. **Use proper formatting**: Match the template's markdown structure exactly (headings, checkboxes, code blocks)
+- **Follow PR template**: When submitting pull requests, follow the template in `.github/pull_request_template.md` to ensure complete context and documentation.
 
 ## Development Commands
 

electron-builder.yml

@@ -134,108 +134,56 @@ artifactBuildCompleted: scripts/artifact-build-completed.js
 releaseInfo:
   releaseNotes: |
     <!--LANG:en-->
-    A New Era of Intelligence with Cherry Studio 1.7.1
+    What's New in v1.7.0-rc.3
 
-    Today we're releasing Cherry Studio 1.7.1 — our most ambitious update yet, introducing Agent: autonomous AI that thinks, plans, and acts.
+    ✨ New Features:
+    - Provider: Added Silicon provider support for Anthropic API compatibility
+    - Provider: AIHubMix support for nano banana
 
-    For years, AI assistants have been reactive — waiting for your commands, responding to your questions. With Agent, we're changing that. Now, AI can truly work alongside you: understanding complex goals, breaking them into steps, and executing them independently.
+    🐛 Bug Fixes:
+    - i18n: Clean up translation tags and untranslated strings
+    - Provider: Fixed Silicon provider code list
+    - Provider: Fixed Poe API reasoning parameters for GPT-5 and reasoning models
+    - Provider: Fixed duplicate /v1 in Anthropic API endpoints
+    - Provider: Fixed Azure provider handling in AI SDK integration
+    - Models: Added Claude Opus 4.5 pattern to THINKING_TOKEN_MAP
+    - Models: Improved Gemini reasoning and message handling
+    - Models: Fixed custom parameters for Gemini models
+    - Models: Fixed qwen-mt-flash text delta support
+    - Models: Fixed Groq verbosity setting
+    - UI: Fixed quota display and quota tips
+    - UI: Fixed web search button condition
+    - Settings: Fixed updateAssistantPreset reducer to properly update preset
+    - Settings: Respect enableMaxTokens setting when maxTokens is not configured
+    - SDK: Fixed header merging logic in AI SDK
 
-    This is what we've been building toward. And it's just the beginning.
-
-    🤖 Meet Agent
-    Imagine having a brilliant colleague who never sleeps. Give Agent a goal — write a report, analyze data, refactor code — and watch it work. It reasons through problems, breaks them into steps, calls the right tools, and adapts when things change.
-
-    - **Think → Plan → Act**: From goal to execution, fully autonomous
-    - **Deep Reasoning**: Multi-turn thinking that solves real problems
-    - **Tool Mastery**: File operations, web search, code execution, and more
-    - **Skill Plugins**: Extend with custom commands and capabilities
-    - **You Stay in Control**: Real-time approval for sensitive actions
-    - **Full Visibility**: Every thought, every decision, fully transparent
-
-    🌐 Expanding Ecosystem
-    - **New Providers**: HuggingFace, Mistral, CherryIN, AI Gateway, Intel OVMS, Didi MCP
-    - **New Models**: Claude 4.5 Haiku, DeepSeek v3.2, GLM-4.6, Doubao, Ling series
-    - **MCP Integration**: Alibaba Cloud, ModelScope, Higress, MCP.so, TokenFlux and more
-
-    📚 Smarter Knowledge Base
-    - **OpenMinerU**: Self-hosted document processing
-    - **Full-Text Search**: Find anything instantly across your notes
-    - **Enhanced Tool Selection**: Smarter configuration for better AI assistance
-
-    📝 Notes, Reimagined
-    - Full-text search with highlighted results
-    - AI-powered smart rename
-    - Export as image
-    - Auto-wrap for tables
-
-    🖼️ Image & OCR
-    - Intel OVMS painting capabilities
-    - Intel OpenVINO NPU-accelerated OCR
-
-    🌍 Now in 10+ Languages
-    - Added German support
-    - Enhanced internationalization
-
-    ⚡ Faster & More Polished
-    - Electron 38 upgrade
-    - New MCP management interface
-    - Dozens of UI refinements
-
-    ❤️ Fully Open Source
-    Commercial restrictions removed. Cherry Studio now follows standard AGPL v3 — free for teams of any size.
-
-    The Agent Era is here. We can't wait to see what you'll create.
+    ⚡ Improvements:
+    - SDK: Upgraded @anthropic-ai/claude-agent-sdk to 0.1.53
 
     <!--LANG:zh-CN-->
-    Cherry Studio 1.7.1：开启智能新纪元
+    v1.7.0-rc.3 更新内容
 
-    今天，我们正式发布 Cherry Studio 1.7.1 —— 迄今最具雄心的版本，带来全新的 Agent：能够自主思考、规划和行动的 AI。
+    ✨ 新功能:
+    - 提供商：新增 Silicon 提供商对 Anthropic API 的兼容性支持
+    - 提供商：AIHubMix 支持 nano banana
 
-    多年来，AI 助手一直是被动的——等待你的指令，回应你的问题。Agent 改变了这一切。现在，AI 能够真正与你并肩工作：理解复杂目标，将其拆解为步骤，并独立执行。
+    🐛 问题修复:
+    - 国际化：清理翻译标签和未翻译字符串
+    - 提供商：修复 Silicon 提供商代码列表
+    - 提供商：修复 Poe API 对 GPT-5 和推理模型的推理参数
+    - 提供商：修复 Anthropic API 端点重复 /v1 问题
+    - 提供商：修复 Azure 提供商在 AI SDK 集成中的处理
+    - 模型：Claude Opus 4.5 添加到 THINKING_TOKEN_MAP
+    - 模型：改进 Gemini 推理和消息处理
+    - 模型：修复 Gemini 模型自定义参数
+    - 模型：修复 qwen-mt-flash text delta 支持
+    - 模型：修复 Groq verbosity 设置
+    - 界面：修复配额显示和配额提示
+    - 界面：修复 Web 搜索按钮条件
+    - 设置：修复 updateAssistantPreset reducer 正确更新 preset
+    - 设置：尊重 enableMaxTokens 设置
+    - SDK：修复 AI SDK 中 header 合并逻辑
 
-    这是我们一直在构建的未来。而这，仅仅是开始。
-
-    🤖 认识 Agent
-    想象一位永不疲倦的得力伙伴。给 Agent 一个目标——撰写报告、分析数据、重构代码——然后看它工作。它会推理问题、拆解步骤、调用工具，并在情况变化时灵活应对。
-
-    - **思考 → 规划 → 行动**：从目标到执行，全程自主
-    - **深度推理**：多轮思考，解决真实问题
-    - **工具大师**：文件操作、网络搜索、代码执行，样样精通
-    - **技能插件**：自定义命令，无限扩展
-    - **你掌控全局**：敏感操作，实时审批
-    - **完全透明**：每一步思考，每一个决策，清晰可见
-
-    🌐 生态持续壮大
-    - **新增服务商**：Hugging Face、Mistral、Perplexity、SophNet、AI Gateway、Cerebras AI
-    - **新增模型**：Gemini 3、Gemini 3 Pro（支持图像预览）、GPT-5.1、Claude Opus 4.5
-    - **MCP 集成**：百炼、魔搭、Higress、MCP.so、TokenFlux 等平台
-
-    📚 更智能的知识库
-    - **OpenMinerU**：本地自部署文档处理
-    - **全文搜索**：笔记内容一搜即达
-    - **增强工具选择**：更智能的配置，更好的 AI 协助
-
-    📝 笔记，焕然一新
-    - 全文搜索，结果高亮
-    - AI 智能重命名
-    - 导出为图片
-    - 表格自动换行
-
-    🖼️ 图像与 OCR
-    - Intel OVMS 绘图能力
-    - Intel OpenVINO NPU 加速 OCR
-
-    🌍 支持 10+ 种语言
-    - 新增德语支持
-    - 全面增强国际化
-
-    ⚡ 更快、更精致
-    - 升级 Electron 38
-    - 新的 MCP 管理界面
-    - 数十处 UI 细节打磨
-
-    ❤️ 完全开源
-    商用限制已移除。Cherry Studio 现遵循标准 AGPL v3 协议——任意规模团队均可自由使用。
-
-    Agent 纪元已至。期待你的创造。
+    ⚡ 改进:
+    - SDK：升级 @anthropic-ai/claude-agent-sdk 到 0.1.53
     <!--LANG:END-->

eslint.config.mjs

@@ -58,7 +58,6 @@ export default defineConfig([
       'dist/**',
       'out/**',
       'local/**',
-      'tests/**',
       '.yarn/**',
       '.gitignore',
       'scripts/cloudflare-worker.js',

package.json

@@ -1,6 +1,6 @@
 {
   "name": "CherryStudio",
-  "version": "1.7.1",
+  "version": "1.7.0-rc.3",
   "private": true,
   "description": "A powerful AI assistant for producer.",
   "main": "./out/main/index.js",
@@ -62,7 +62,6 @@
     "test": "vitest run --silent",
     "test:main": "vitest run --project main",
     "test:renderer": "vitest run --project renderer",
-    "test:aicore": "vitest run --project aiCore",
     "test:update": "yarn test:renderer --update",
     "test:coverage": "vitest run --coverage --silent",
     "test:ui": "vitest --ui",
@@ -165,7 +164,7 @@
     "@modelcontextprotocol/sdk": "^1.17.5",
     "@mozilla/readability": "^0.6.0",
     "@notionhq/client": "^2.2.15",
-    "@openrouter/ai-sdk-provider": "^1.2.8",
+    "@openrouter/ai-sdk-provider": "^1.2.5",
     "@opentelemetry/api": "^1.9.0",
     "@opentelemetry/core": "2.0.0",
     "@opentelemetry/exporter-trace-otlp-http": "^0.200.0",
@@ -173,7 +172,7 @@
     "@opentelemetry/sdk-trace-node": "^2.0.0",
     "@opentelemetry/sdk-trace-web": "^2.0.0",
     "@opeoginni/github-copilot-openai-compatible": "^0.1.21",
-    "@playwright/test": "^1.55.1",
+    "@playwright/test": "^1.52.0",
     "@radix-ui/react-context-menu": "^2.2.16",
     "@reduxjs/toolkit": "^2.2.5",
     "@shikijs/markdown-it": "^3.12.0",
@@ -322,6 +321,7 @@
     "p-queue": "^8.1.0",
     "pdf-lib": "^1.17.1",
     "pdf-parse": "^1.1.1",
+    "playwright": "^1.55.1",
     "proxy-agent": "^6.5.0",
     "react": "^19.2.0",
     "react-dom": "^19.2.0",

packages/ai-sdk-provider/src/cherryin-provider.ts

@@ -69,7 +69,6 @@ export interface CherryInProviderSettings {
   headers?: HeadersInput
   /**
    * Optional endpoint type to distinguish different endpoint behaviors.
-   * "image-generation" is also openai endpoint, but specifically for image generation.
    */
   endpointType?: 'openai' | 'openai-response' | 'anthropic' | 'gemini' | 'image-generation' | 'jina-rerank'
 }

packages/aiCore/src/__tests__/fixtures/mock-responses.ts

@@ -3,13 +3,12 @@
  * Provides realistic mock responses for all provider types
  */
 
-import type { ModelMessage, Tool } from 'ai'
-import { jsonSchema } from 'ai'
+import { jsonSchema, type ModelMessage, type Tool } from 'ai'
 
 /**
  * Standard test messages for all scenarios
  */
-export const testMessages: Record<string, ModelMessage[]> = {
+export const testMessages = {
   simple: [{ role: 'user' as const, content: 'Hello, how are you?' }],
 
   conversation: [
@@ -46,7 +45,7 @@ export const testMessages: Record<string, ModelMessage[]> = {
     { role: 'assistant' as const, content: '15 * 23 = 345' },
     { role: 'user' as const, content: 'Now divide that by 5' }
   ]
-}
+} satisfies Record<string, ModelMessage[]>
 
 /**
  * Standard test tools for tool calling scenarios
@@ -139,17 +138,68 @@ export const testTools: Record<string, Tool> = {
   }
 }
 
+/**
+ * Mock streaming chunks for different providers
+ */
+export const mockStreamingChunks = {
+  text: [
+    { type: 'text-delta' as const, textDelta: 'Hello' },
+    { type: 'text-delta' as const, textDelta: ', ' },
+    { type: 'text-delta' as const, textDelta: 'this ' },
+    { type: 'text-delta' as const, textDelta: 'is ' },
+    { type: 'text-delta' as const, textDelta: 'a ' },
+    { type: 'text-delta' as const, textDelta: 'test.' }
+  ],
+
+  withToolCall: [
+    { type: 'text-delta' as const, textDelta: 'Let me check the weather for you.' },
+    {
+      type: 'tool-call-delta' as const,
+      toolCallType: 'function' as const,
+      toolCallId: 'call_123',
+      toolName: 'getWeather',
+      argsTextDelta: '{"location":'
+    },
+    {
+      type: 'tool-call-delta' as const,
+      toolCallType: 'function' as const,
+      toolCallId: 'call_123',
+      toolName: 'getWeather',
+      argsTextDelta: ' "San Francisco, CA"}'
+    },
+    {
+      type: 'tool-call' as const,
+      toolCallType: 'function' as const,
+      toolCallId: 'call_123',
+      toolName: 'getWeather',
+      args: { location: 'San Francisco, CA' }
+    }
+  ],
+
+  withFinish: [
+    { type: 'text-delta' as const, textDelta: 'Complete response.' },
+    {
+      type: 'finish' as const,
+      finishReason: 'stop' as const,
+      usage: {
+        promptTokens: 10,
+        completionTokens: 5,
+        totalTokens: 15
+      }
+    }
+  ]
+}
+
 /**
  * Mock complete responses for non-streaming scenarios
- * Note: AI SDK v5 uses inputTokens/outputTokens instead of promptTokens/completionTokens
  */
 export const mockCompleteResponses = {
   simple: {
     text: 'This is a simple response.',
     finishReason: 'stop' as const,
     usage: {
-      inputTokens: 15,
-      outputTokens: 8,
+      promptTokens: 15,
+      completionTokens: 8,
       totalTokens: 23
     }
   },
@@ -165,8 +215,8 @@ export const mockCompleteResponses = {
     ],
     finishReason: 'tool-calls' as const,
     usage: {
-      inputTokens: 25,
-      outputTokens: 12,
+      promptTokens: 25,
+      completionTokens: 12,
       totalTokens: 37
     }
   },
@@ -175,15 +225,14 @@ export const mockCompleteResponses = {
     text: 'Response with warnings.',
     finishReason: 'stop' as const,
     usage: {
-      inputTokens: 10,
-      outputTokens: 5,
+      promptTokens: 10,
+      completionTokens: 5,
       totalTokens: 15
     },
     warnings: [
       {
         type: 'unsupported-setting' as const,
-        setting: 'temperature',
-        details: 'Temperature parameter not supported for this model'
+        message: 'Temperature parameter not supported for this model'
       }
     ]
   }
@@ -236,3 +285,47 @@ export const mockImageResponses = {
     warnings: []
   }
 }
+
+/**
+ * Mock error responses
+ */
+export const mockErrors = {
+  invalidApiKey: {
+    name: 'APIError',
+    message: 'Invalid API key provided',
+    statusCode: 401
+  },
+
+  rateLimitExceeded: {
+    name: 'RateLimitError',
+    message: 'Rate limit exceeded. Please try again later.',
+    statusCode: 429,
+    headers: {
+      'retry-after': '60'
+    }
+  },
+
+  modelNotFound: {
+    name: 'ModelNotFoundError',
+    message: 'The requested model was not found',
+    statusCode: 404
+  },
+
+  contextLengthExceeded: {
+    name: 'ContextLengthError',
+    message: "This model's maximum context length is 4096 tokens",
+    statusCode: 400
+  },
+
+  timeout: {
+    name: 'TimeoutError',
+    message: 'Request timed out after 30000ms',
+    code: 'ETIMEDOUT'
+  },
+
+  networkError: {
+    name: 'NetworkError',
+    message: 'Network connection failed',
+    code: 'ECONNREFUSED'
+  }
+}

packages/aiCore/src/__tests__/mocks/ai-sdk-provider.ts

@@ -1,35 +0,0 @@
-/**
- * Mock for @cherrystudio/ai-sdk-provider
- * This mock is used in tests to avoid importing the actual package
- */
-
-export type CherryInProviderSettings = {
-  apiKey?: string
-  baseURL?: string
-}
-
-// oxlint-disable-next-line no-unused-vars
-export const createCherryIn = (_options?: CherryInProviderSettings) => ({
-  // oxlint-disable-next-line no-unused-vars
-  languageModel: (_modelId: string) => ({
-    specificationVersion: 'v1',
-    provider: 'cherryin',
-    modelId: 'mock-model',
-    doGenerate: async () => ({ text: 'mock response' }),
-    doStream: async () => ({ stream: (async function* () {})() })
-  }),
-  // oxlint-disable-next-line no-unused-vars
-  chat: (_modelId: string) => ({
-    specificationVersion: 'v1',
-    provider: 'cherryin-chat',
-    modelId: 'mock-model',
-    doGenerate: async () => ({ text: 'mock response' }),
-    doStream: async () => ({ stream: (async function* () {})() })
-  }),
-  // oxlint-disable-next-line no-unused-vars
-  textEmbeddingModel: (_modelId: string) => ({
-    specificationVersion: 'v1',
-    provider: 'cherryin',
-    modelId: 'mock-embedding-model'
-  })
-})

packages/aiCore/src/__tests__/setup.ts

@@ -1,9 +0,0 @@
-/**
- * Vitest Setup File
- * Global test configuration and mocks for @cherrystudio/ai-core package
- */
-
-// Mock Vite SSR helper to avoid Node environment errors
-;(globalThis as any).__vite_ssr_exportName__ = (_name: string, value: any) => value
-
-// Note: @cherrystudio/ai-sdk-provider is mocked via alias in vitest.config.ts

packages/aiCore/src/core/options/__tests__/factory.test.ts

@@ -1,109 +0,0 @@
-import { describe, expect, it } from 'vitest'
-
-import { createOpenAIOptions, createOpenRouterOptions, mergeProviderOptions } from '../factory'
-
-describe('mergeProviderOptions', () => {
-  it('deep merges provider options for the same provider', () => {
-    const reasoningOptions = createOpenRouterOptions({
-      reasoning: {
-        enabled: true,
-        effort: 'medium'
-      }
-    })
-    const webSearchOptions = createOpenRouterOptions({
-      plugins: [{ id: 'web', max_results: 5 }]
-    })
-
-    const merged = mergeProviderOptions(reasoningOptions, webSearchOptions)
-
-    expect(merged.openrouter).toEqual({
-      reasoning: {
-        enabled: true,
-        effort: 'medium'
-      },
-      plugins: [{ id: 'web', max_results: 5 }]
-    })
-  })
-
-  it('preserves options from other providers while merging', () => {
-    const openRouter = createOpenRouterOptions({
-      reasoning: { enabled: true }
-    })
-    const openAI = createOpenAIOptions({
-      reasoningEffort: 'low'
-    })
-    const merged = mergeProviderOptions(openRouter, openAI)
-
-    expect(merged.openrouter).toEqual({ reasoning: { enabled: true } })
-    expect(merged.openai).toEqual({ reasoningEffort: 'low' })
-  })
-
-  it('overwrites primitive values with later values', () => {
-    const first = createOpenAIOptions({
-      reasoningEffort: 'low',
-      user: 'user-123'
-    })
-    const second = createOpenAIOptions({
-      reasoningEffort: 'high',
-      maxToolCalls: 5
-    })
-
-    const merged = mergeProviderOptions(first, second)
-
-    expect(merged.openai).toEqual({
-      reasoningEffort: 'high', // overwritten by second
-      user: 'user-123', // preserved from first
-      maxToolCalls: 5 // added from second
-    })
-  })
-
-  it('overwrites arrays with later values instead of merging', () => {
-    const first = createOpenRouterOptions({
-      models: ['gpt-4', 'gpt-3.5-turbo']
-    })
-    const second = createOpenRouterOptions({
-      models: ['claude-3-opus', 'claude-3-sonnet']
-    })
-
-    const merged = mergeProviderOptions(first, second)
-
-    // Array is completely replaced, not merged
-    expect(merged.openrouter?.models).toEqual(['claude-3-opus', 'claude-3-sonnet'])
-  })
-
-  it('deeply merges nested objects while overwriting primitives', () => {
-    const first = createOpenRouterOptions({
-      reasoning: {
-        enabled: true,
-        effort: 'low'
-      },
-      user: 'user-123'
-    })
-    const second = createOpenRouterOptions({
-      reasoning: {
-        effort: 'high',
-        max_tokens: 500
-      },
-      user: 'user-456'
-    })
-
-    const merged = mergeProviderOptions(first, second)
-
-    expect(merged.openrouter).toEqual({
-      reasoning: {
-        enabled: true, // preserved from first
-        effort: 'high', // overwritten by second
-        max_tokens: 500 // added from second
-      },
-      user: 'user-456' // overwritten by second
-    })
-  })
-
-  it('replaces arrays instead of merging them', () => {
-    const first = createOpenRouterOptions({ plugins: [{ id: 'old' }] })
-    const second = createOpenRouterOptions({ plugins: [{ id: 'new' }] })
-    const merged = mergeProviderOptions(first, second)
-    // @ts-expect-error type-check for openrouter options is skipped. see function signature of createOpenRouterOptions
-    expect(merged.openrouter?.plugins).toEqual([{ id: 'new' }])
-  })
-})

packages/aiCore/src/core/options/factory.ts

@@ -26,65 +26,13 @@ export function createGenericProviderOptions<T extends string>(
   return { [provider]: options } as Record<T, Record<string, any>>
 }
 
-type PlainObject = Record<string, any>
-
-const isPlainObject = (value: unknown): value is PlainObject => {
-  return typeof value === 'object' && value !== null && !Array.isArray(value)
-}
-
-function deepMergeObjects<T extends PlainObject>(target: T, source: PlainObject): T {
-  const result: PlainObject = { ...target }
-  Object.entries(source).forEach(([key, value]) => {
-    if (isPlainObject(value) && isPlainObject(result[key])) {
-      result[key] = deepMergeObjects(result[key], value)
-    } else {
-      result[key] = value
-    }
-  })
-  return result as T
-}
-
 /**
- * Deep-merge multiple provider-specific options.
- * Nested objects are recursively merged; primitive values are overwritten.
- *
- * When the same key appears in multiple options:
- * - If both values are plain objects: they are deeply merged (recursive merge)
- * - If values are primitives/arrays: the later value overwrites the earlier one
- *
- * @example
- * mergeProviderOptions(
- *   { openrouter: { reasoning: { enabled: true, effort: 'low' }, user: 'user-123' } },
- *   { openrouter: { reasoning: { effort: 'high', max_tokens: 500 }, models: ['gpt-4'] } }
- * )
- * // Result: {
- * //   openrouter: {
- * //     reasoning: { enabled: true, effort: 'high', max_tokens: 500 },
- * //     user: 'user-123',
- * //     models: ['gpt-4']
- * //   }
- * // }
- *
- * @param optionsMap Objects containing options for multiple providers
- * @returns Fully merged TypedProviderOptions
+ * 合并多个供应商的options
+ * @param optionsMap 包含多个供应商选项的对象
+ * @returns 合并后的TypedProviderOptions
  */
 export function mergeProviderOptions(...optionsMap: Partial<TypedProviderOptions>[]): TypedProviderOptions {
-  return optionsMap.reduce<TypedProviderOptions>((acc, options) => {
-    if (!options) {
-      return acc
-    }
-    Object.entries(options).forEach(([providerId, providerOptions]) => {
-      if (!providerOptions) {
-        return
-      }
-      if (acc[providerId]) {
-        acc[providerId] = deepMergeObjects(acc[providerId] as PlainObject, providerOptions as PlainObject)
-      } else {
-        acc[providerId] = providerOptions as any
-      }
-    })
-    return acc
-  }, {} as TypedProviderOptions)
+  return Object.assign({}, ...optionsMap)
 }
 
 /**

packages/aiCore/src/core/providers/__tests__/schemas.test.ts

@@ -19,20 +19,15 @@ describe('Provider Schemas', () => {
       expect(Array.isArray(baseProviders)).toBe(true)
       expect(baseProviders.length).toBeGreaterThan(0)
 
-      // These are the actual base providers defined in schemas.ts
       const expectedIds = [
         'openai',
-        'openai-chat',
+        'openai-responses',
         'openai-compatible',
         'anthropic',
         'google',
         'xai',
         'azure',
-        'azure-responses',
-        'deepseek',
-        'openrouter',
-        'cherryin',
-        'cherryin-chat'
+        'deepseek'
       ]
       const actualIds = baseProviders.map((p) => p.id)
       expectedIds.forEach((id) => {

packages/aiCore/src/core/runtime/__tests__/generateImage.test.ts

@@ -232,13 +232,11 @@ describe('RuntimeExecutor.generateImage', () => {
 
       expect(pluginCallOrder).toEqual(['onRequestStart', 'transformParams', 'transformResult', 'onRequestEnd'])
 
-      // transformParams receives params without model (model is handled separately)
-      // and context with core fields + dynamic fields (requestId, startTime, etc.)
       expect(testPlugin.transformParams).toHaveBeenCalledWith(
-        expect.objectContaining({ prompt: 'A test image' }),
+        { prompt: 'A test image' },
         expect.objectContaining({
           providerId: 'openai',
-          model: 'dall-e-3'
+          modelId: 'dall-e-3'
         })
       )
 
@@ -275,12 +273,11 @@ describe('RuntimeExecutor.generateImage', () => {
 
       await executorWithPlugin.generateImage({ model: 'dall-e-3', prompt: 'A test image' })
 
-      // resolveModel receives model id and context with core fields
       expect(modelResolutionPlugin.resolveModel).toHaveBeenCalledWith(
         'dall-e-3',
         expect.objectContaining({
           providerId: 'openai',
-          model: 'dall-e-3'
+          modelId: 'dall-e-3'
         })
       )
 
@@ -342,11 +339,12 @@ describe('RuntimeExecutor.generateImage', () => {
         .generateImage({ model: 'invalid-model', prompt: 'A test image' })
         .catch((error) => error)
 
-      // Error is thrown from pluginEngine directly as ImageModelResolutionError
-      expect(thrownError).toBeInstanceOf(ImageModelResolutionError)
-      expect(thrownError.message).toContain('Failed to resolve image model: invalid-model')
+      expect(thrownError).toBeInstanceOf(ImageGenerationError)
+      expect(thrownError.message).toContain('Failed to generate image:')
       expect(thrownError.providerId).toBe('openai')
       expect(thrownError.modelId).toBe('invalid-model')
+      expect(thrownError.cause).toBeInstanceOf(ImageModelResolutionError)
+      expect(thrownError.cause.message).toContain('Failed to resolve image model: invalid-model')
     })
 
     it('should handle ImageModelResolutionError without provider', async () => {
@@ -364,9 +362,8 @@ describe('RuntimeExecutor.generateImage', () => {
       const apiError = new Error('API request failed')
       vi.mocked(aiGenerateImage).mockRejectedValue(apiError)
 
-      // Error propagates directly from pluginEngine without wrapping
       await expect(executor.generateImage({ model: 'dall-e-3', prompt: 'A test image' })).rejects.toThrow(
-        'API request failed'
+        'Failed to generate image:'
       )
     })
 
@@ -379,9 +376,8 @@ describe('RuntimeExecutor.generateImage', () => {
       vi.mocked(aiGenerateImage).mockRejectedValue(noImageError)
       vi.mocked(NoImageGeneratedError.isInstance).mockReturnValue(true)
 
-      // Error propagates directly from pluginEngine
       await expect(executor.generateImage({ model: 'dall-e-3', prompt: 'A test image' })).rejects.toThrow(
-        'No image generated'
+        'Failed to generate image:'
       )
     })
 
@@ -402,17 +398,15 @@ describe('RuntimeExecutor.generateImage', () => {
         [errorPlugin]
       )
 
-      // Error propagates directly from pluginEngine
       await expect(executorWithPlugin.generateImage({ model: 'dall-e-3', prompt: 'A test image' })).rejects.toThrow(
-        'Generation failed'
+        'Failed to generate image:'
       )
 
-      // onError receives the original error and context with core fields
       expect(errorPlugin.onError).toHaveBeenCalledWith(
         error,
         expect.objectContaining({
           providerId: 'openai',
-          model: 'dall-e-3'
+          modelId: 'dall-e-3'
         })
       )
     })
@@ -425,10 +419,9 @@ describe('RuntimeExecutor.generateImage', () => {
       const abortController = new AbortController()
       setTimeout(() => abortController.abort(), 10)
 
-      // Error propagates directly from pluginEngine
       await expect(
         executor.generateImage({ model: 'dall-e-3', prompt: 'A test image', abortSignal: abortController.signal })
-      ).rejects.toThrow('Operation was aborted')
+      ).rejects.toThrow('Failed to generate image:')
     })
   })
 

packages/aiCore/src/core/runtime/__tests__/generateText.test.ts

@@ -17,14 +17,10 @@ import type { AiPlugin } from '../../plugins'
 import { globalRegistryManagement } from '../../providers/RegistryManagement'
 import { RuntimeExecutor } from '../executor'
 
-// Mock AI SDK - use importOriginal to keep jsonSchema and other non-mocked exports
-vi.mock('ai', async (importOriginal) => {
-  const actual = (await importOriginal()) as Record<string, unknown>
-  return {
-    ...actual,
-    generateText: vi.fn()
-  }
-})
+// Mock AI SDK
+vi.mock('ai', () => ({
+  generateText: vi.fn()
+}))
 
 vi.mock('../../providers/RegistryManagement', () => ({
   globalRegistryManagement: {
@@ -413,12 +409,11 @@ describe('RuntimeExecutor.generateText', () => {
         })
       ).rejects.toThrow('Generation failed')
 
-      // onError receives the original error and context with core fields
       expect(errorPlugin.onError).toHaveBeenCalledWith(
         error,
         expect.objectContaining({
           providerId: 'openai',
-          model: 'gpt-4'
+          modelId: 'gpt-4'
         })
       )
     })

packages/aiCore/src/core/runtime/__tests__/streamText.test.ts

@@ -11,14 +11,10 @@ import type { AiPlugin } from '../../plugins'
 import { globalRegistryManagement } from '../../providers/RegistryManagement'
 import { RuntimeExecutor } from '../executor'
 
-// Mock AI SDK - use importOriginal to keep jsonSchema and other non-mocked exports
-vi.mock('ai', async (importOriginal) => {
-  const actual = (await importOriginal()) as Record<string, unknown>
-  return {
-    ...actual,
-    streamText: vi.fn()
-  }
-})
+// Mock AI SDK
+vi.mock('ai', () => ({
+  streamText: vi.fn()
+}))
 
 vi.mock('../../providers/RegistryManagement', () => ({
   globalRegistryManagement: {
@@ -157,7 +153,7 @@ describe('RuntimeExecutor.streamText', () => {
   describe('Max Tokens Parameter', () => {
     const maxTokensValues = [10, 50, 100, 500, 1000, 2000, 4000]
 
-    it.each(maxTokensValues)('should support maxOutputTokens=%s', async (maxOutputTokens) => {
+    it.each(maxTokensValues)('should support maxTokens=%s', async (maxTokens) => {
       const mockStream = {
         textStream: (async function* () {
           yield 'Response'
@@ -172,13 +168,12 @@ describe('RuntimeExecutor.streamText', () => {
       await executor.streamText({
         model: 'gpt-4',
         messages: testMessages.simple,
-        maxOutputTokens
+        maxOutputTokens: maxTokens
       })
 
-      // Parameters are passed through without transformation
       expect(streamText).toHaveBeenCalledWith(
         expect.objectContaining({
-          maxOutputTokens
+          maxTokens
         })
       )
     })
@@ -518,12 +513,11 @@ describe('RuntimeExecutor.streamText', () => {
         })
       ).rejects.toThrow('Stream error')
 
-      // onError receives the original error and context with core fields
       expect(errorPlugin.onError).toHaveBeenCalledWith(
         error,
         expect.objectContaining({
           providerId: 'openai',
-          model: 'gpt-4'
+          modelId: 'gpt-4'
         })
       )
     })

packages/aiCore/vitest.config.ts

@@ -1,20 +1,12 @@
-import path from 'node:path'
-import { fileURLToPath } from 'node:url'
-
 import { defineConfig } from 'vitest/config'
 
-const __dirname = path.dirname(fileURLToPath(import.meta.url))
-
 export default defineConfig({
   test: {
-    globals: true,
-    setupFiles: [path.resolve(__dirname, './src/__tests__/setup.ts')]
+    globals: true
   },
   resolve: {
     alias: {
-      '@': path.resolve(__dirname, './src'),
-      // Mock external packages that may not be available in test environment
-      '@cherrystudio/ai-sdk-provider': path.resolve(__dirname, './src/__tests__/mocks/ai-sdk-provider.ts')
+      '@': './src'
     }
   },
   esbuild: {

playwright.config.ts

@@ -1,64 +1,42 @@
-import { defineConfig } from '@playwright/test'
+import { defineConfig, devices } from '@playwright/test'
 
 /**
- * Playwright configuration for Electron e2e testing.
- * See https://playwright.dev/docs/test-configuration
+ * See https://playwright.dev/docs/test-configuration.
  */
 export default defineConfig({
-  // Look for test files in the specs directory
-  testDir: './tests/e2e/specs',
-
-  // Global timeout for each test
-  timeout: 60000,
-
-  // Assertion timeout
-  expect: {
-    timeout: 10000
-  },
-
-  // Electron apps should run tests sequentially to avoid conflicts
-  fullyParallel: false,
-  workers: 1,
-
-  // Fail the build on CI if you accidentally left test.only in the source code
+  // Look for test files, relative to this configuration file.
+  testDir: './tests/e2e',
+  /* Run tests in files in parallel */
+  fullyParallel: true,
+  /* Fail the build on CI if you accidentally left test.only in the source code. */
   forbidOnly: !!process.env.CI,
-
-  // Retry on CI only
+  /* Retry on CI only */
   retries: process.env.CI ? 2 : 0,
-
-  // Reporter configuration
-  reporter: [['html', { outputFolder: 'playwright-report' }], ['list']],
-
-  // Global setup and teardown
-  globalSetup: './tests/e2e/global-setup.ts',
-  globalTeardown: './tests/e2e/global-teardown.ts',
-
-  // Output directory for test artifacts
-  outputDir: './test-results',
-
-  // Shared settings for all tests
+  /* Opt out of parallel tests on CI. */
+  workers: process.env.CI ? 1 : undefined,
+  /* Reporter to use. See https://playwright.dev/docs/test-reporters */
+  reporter: 'html',
+  /* Shared settings for all the projects below. See https://playwright.dev/docs/api/class-testoptions. */
   use: {
-    // Collect trace when retrying the failed test
-    trace: 'retain-on-failure',
+    /* Base URL to use in actions like `await page.goto('/')`. */
+    // baseURL: 'http://localhost:3000',
 
-    // Take screenshot only on failure
-    screenshot: 'only-on-failure',
-
-    // Record video only on failure
-    video: 'retain-on-failure',
-
-    // Action timeout
-    actionTimeout: 15000,
-
-    // Navigation timeout
-    navigationTimeout: 30000
+    /* Collect trace when retrying the failed test. See https://playwright.dev/docs/trace-viewer */
+    trace: 'on-first-retry'
   },
 
-  // Single project for Electron testing
+  /* Configure projects for major browsers */
   projects: [
     {
-      name: 'electron',
-      testMatch: '**/*.spec.ts'
+      name: 'chromium',
+      use: { ...devices['Desktop Chrome'] }
     }
   ]
+
+  /* Run your local dev server before starting the tests */
+  // webServer: {
+  //   command: 'npm run start',
+  //   url: 'http://localhost:3000',
+  //   reuseExistingServer: !process.env.CI,
+  // },
 })

src/main/services/CodeToolsService.ts

@@ -548,17 +548,6 @@ class CodeToolsService {
     logger.debug(`Environment variables:`, Object.keys(env))
     logger.debug(`Options:`, options)
 
-    // Validate directory exists before proceeding
-    if (!directory || !fs.existsSync(directory)) {
-      const errorMessage = `Directory does not exist: ${directory}`
-      logger.error(errorMessage)
-      return {
-        success: false,
-        message: errorMessage,
-        command: ''
-      }
-    }
-
     const packageName = await this.getPackageName(cliTool)
     const bunPath = await this.getBunPath()
     const executableName = await this.getCliExecutableName(cliTool)
@@ -720,7 +709,6 @@ class CodeToolsService {
         // Build bat file content, including debug information
         const batContent = [
           '@echo off',
-          'chcp 65001 >nul 2>&1', // Switch to UTF-8 code page for international path support
           `title ${cliTool} - Cherry Studio`, // Set window title in bat file
           'echo ================================================',
           'echo Cherry Studio CLI Tool Launcher',

src/main/services/MCPService.ts

@@ -620,7 +620,7 @@ class McpService {
       tools.map((tool: SDKTool) => {
         const serverTool: MCPTool = {
           ...tool,
-          id: buildFunctionCallToolName(server.name, tool.name, server.id),
+          id: buildFunctionCallToolName(server.name, tool.name),
           serverId: server.id,
           serverName: server.name,
           type: 'mcp'

src/main/services/agents/services/claudecode/index.ts

@@ -1,7 +1,6 @@
 // src/main/services/agents/services/claudecode/index.ts
 import { EventEmitter } from 'node:events'
 import { createRequire } from 'node:module'
-import path from 'node:path'
 
 import type {
   CanUseTool,
@@ -122,11 +121,7 @@ class ClaudeCodeService implements AgentServiceInterface {
       // TODO: support set small model in UI
       ANTHROPIC_DEFAULT_HAIKU_MODEL: modelInfo.modelId,
       ELECTRON_RUN_AS_NODE: '1',
-      ELECTRON_NO_ATTACH_CONSOLE: '1',
-      // Set CLAUDE_CONFIG_DIR to app's userData directory to avoid path encoding issues
-      // on Windows when the username contains non-ASCII characters (e.g., Chinese characters)
-      // This prevents the SDK from using the user's home directory which may have encoding problems
-      CLAUDE_CONFIG_DIR: path.join(app.getPath('userData'), '.claude')
+      ELECTRON_NO_ATTACH_CONSOLE: '1'
     }
 
     const errorChunks: string[] = []

src/main/utils/__tests__/mcp.test.ts

@@ -1,196 +0,0 @@
-import { describe, expect, it } from 'vitest'
-
-import { buildFunctionCallToolName } from '../mcp'
-
-describe('buildFunctionCallToolName', () => {
-  describe('basic functionality', () => {
-    it('should combine server name and tool name', () => {
-      const result = buildFunctionCallToolName('github', 'search_issues')
-      expect(result).toContain('github')
-      expect(result).toContain('search')
-    })
-
-    it('should sanitize names by replacing dashes with underscores', () => {
-      const result = buildFunctionCallToolName('my-server', 'my-tool')
-      // Input dashes are replaced, but the separator between server and tool is a dash
-      expect(result).toBe('my_serv-my_tool')
-      expect(result).toContain('_')
-    })
-
-    it('should handle empty server names gracefully', () => {
-      const result = buildFunctionCallToolName('', 'tool')
-      expect(result).toBeTruthy()
-    })
-  })
-
-  describe('uniqueness with serverId', () => {
-    it('should generate different IDs for same server name but different serverIds', () => {
-      const serverId1 = 'server-id-123456'
-      const serverId2 = 'server-id-789012'
-      const serverName = 'github'
-      const toolName = 'search_repos'
-
-      const result1 = buildFunctionCallToolName(serverName, toolName, serverId1)
-      const result2 = buildFunctionCallToolName(serverName, toolName, serverId2)
-
-      expect(result1).not.toBe(result2)
-      expect(result1).toContain('123456')
-      expect(result2).toContain('789012')
-    })
-
-    it('should generate same ID when serverId is not provided', () => {
-      const serverName = 'github'
-      const toolName = 'search_repos'
-
-      const result1 = buildFunctionCallToolName(serverName, toolName)
-      const result2 = buildFunctionCallToolName(serverName, toolName)
-
-      expect(result1).toBe(result2)
-    })
-
-    it('should include serverId suffix when provided', () => {
-      const serverId = 'abc123def456'
-      const result = buildFunctionCallToolName('server', 'tool', serverId)
-
-      // Should include last 6 chars of serverId
-      expect(result).toContain('ef456')
-    })
-  })
-
-  describe('character sanitization', () => {
-    it('should replace invalid characters with underscores', () => {
-      const result = buildFunctionCallToolName('test@server', 'tool#name')
-      expect(result).not.toMatch(/[@#]/)
-      expect(result).toMatch(/^[a-zA-Z0-9_-]+$/)
-    })
-
-    it('should ensure name starts with a letter', () => {
-      const result = buildFunctionCallToolName('123server', '456tool')
-      expect(result).toMatch(/^[a-zA-Z]/)
-    })
-
-    it('should handle consecutive underscores/dashes', () => {
-      const result = buildFunctionCallToolName('my--server', 'my__tool')
-      expect(result).not.toMatch(/[_-]{2,}/)
-    })
-  })
-
-  describe('length constraints', () => {
-    it('should truncate names longer than 63 characters', () => {
-      const longServerName = 'a'.repeat(50)
-      const longToolName = 'b'.repeat(50)
-      const result = buildFunctionCallToolName(longServerName, longToolName, 'id123456')
-
-      expect(result.length).toBeLessThanOrEqual(63)
-    })
-
-    it('should not end with underscore or dash after truncation', () => {
-      const longServerName = 'a'.repeat(50)
-      const longToolName = 'b'.repeat(50)
-      const result = buildFunctionCallToolName(longServerName, longToolName, 'id123456')
-
-      expect(result).not.toMatch(/[_-]$/)
-    })
-
-    it('should preserve serverId suffix even with long server/tool names', () => {
-      const longServerName = 'a'.repeat(50)
-      const longToolName = 'b'.repeat(50)
-      const serverId = 'server-id-xyz789'
-
-      const result = buildFunctionCallToolName(longServerName, longToolName, serverId)
-
-      // The suffix should be preserved and not truncated
-      expect(result).toContain('xyz789')
-      expect(result.length).toBeLessThanOrEqual(63)
-    })
-
-    it('should ensure two long-named servers with different IDs produce different results', () => {
-      const longServerName = 'a'.repeat(50)
-      const longToolName = 'b'.repeat(50)
-      const serverId1 = 'server-id-abc123'
-      const serverId2 = 'server-id-def456'
-
-      const result1 = buildFunctionCallToolName(longServerName, longToolName, serverId1)
-      const result2 = buildFunctionCallToolName(longServerName, longToolName, serverId2)
-
-      // Both should be within limit
-      expect(result1.length).toBeLessThanOrEqual(63)
-      expect(result2.length).toBeLessThanOrEqual(63)
-
-      // They should be different due to preserved suffix
-      expect(result1).not.toBe(result2)
-    })
-  })
-
-  describe('edge cases with serverId', () => {
-    it('should handle serverId with only non-alphanumeric characters', () => {
-      const serverId = '------' // All dashes
-      const result = buildFunctionCallToolName('server', 'tool', serverId)
-
-      // Should still produce a valid unique suffix via fallback hash
-      expect(result).toBeTruthy()
-      expect(result.length).toBeLessThanOrEqual(63)
-      expect(result).toMatch(/^[a-zA-Z][a-zA-Z0-9_-]*$/)
-      // Should have a suffix (underscore followed by something)
-      expect(result).toMatch(/_[a-z0-9]+$/)
-    })
-
-    it('should produce different results for different non-alphanumeric serverIds', () => {
-      const serverId1 = '------'
-      const serverId2 = '!!!!!!'
-
-      const result1 = buildFunctionCallToolName('server', 'tool', serverId1)
-      const result2 = buildFunctionCallToolName('server', 'tool', serverId2)
-
-      // Should be different because the hash fallback produces different values
-      expect(result1).not.toBe(result2)
-    })
-
-    it('should handle empty string serverId differently from undefined', () => {
-      const resultWithEmpty = buildFunctionCallToolName('server', 'tool', '')
-      const resultWithUndefined = buildFunctionCallToolName('server', 'tool', undefined)
-
-      // Empty string is falsy, so both should behave the same (no suffix)
-      expect(resultWithEmpty).toBe(resultWithUndefined)
-    })
-
-    it('should handle serverId with mixed alphanumeric and special chars', () => {
-      const serverId = 'ab@#cd' // Mixed chars, last 6 chars contain some alphanumeric
-      const result = buildFunctionCallToolName('server', 'tool', serverId)
-
-      // Should extract alphanumeric chars: 'abcd' from 'ab@#cd'
-      expect(result).toContain('abcd')
-    })
-  })
-
-  describe('real-world scenarios', () => {
-    it('should handle GitHub MCP server instances correctly', () => {
-      const serverName = 'github'
-      const toolName = 'search_repositories'
-
-      const githubComId = 'server-github-com-abc123'
-      const gheId = 'server-ghe-internal-xyz789'
-
-      const tool1 = buildFunctionCallToolName(serverName, toolName, githubComId)
-      const tool2 = buildFunctionCallToolName(serverName, toolName, gheId)
-
-      // Should be different
-      expect(tool1).not.toBe(tool2)
-
-      // Both should be valid identifiers
-      expect(tool1).toMatch(/^[a-zA-Z][a-zA-Z0-9_-]*$/)
-      expect(tool2).toMatch(/^[a-zA-Z][a-zA-Z0-9_-]*$/)
-
-      // Both should be <= 63 chars
-      expect(tool1.length).toBeLessThanOrEqual(63)
-      expect(tool2.length).toBeLessThanOrEqual(63)
-    })
-
-    it('should handle tool names that already include server name prefix', () => {
-      const result = buildFunctionCallToolName('github', 'github_search_repos')
-      expect(result).toBeTruthy()
-      // Should not double the server name
-      expect(result.split('github').length - 1).toBeLessThanOrEqual(2)
-    })
-  })
-})

src/main/utils/mcp.ts

@@ -1,25 +1,7 @@
-export function buildFunctionCallToolName(serverName: string, toolName: string, serverId?: string) {
+export function buildFunctionCallToolName(serverName: string, toolName: string) {
   const sanitizedServer = serverName.trim().replace(/-/g, '_')
   const sanitizedTool = toolName.trim().replace(/-/g, '_')
 
-  // Calculate suffix first to reserve space for it
-  // Suffix format: "_" + 6 alphanumeric chars = 7 chars total
-  let serverIdSuffix = ''
-  if (serverId) {
-    // Take the last 6 characters of the serverId for brevity
-    serverIdSuffix = serverId.slice(-6).replace(/[^a-zA-Z0-9]/g, '')
-
-    // Fallback: if suffix becomes empty (all non-alphanumeric chars), use a simple hash
-    if (!serverIdSuffix) {
-      const hash = serverId.split('').reduce((acc, char) => acc + char.charCodeAt(0), 0)
-      serverIdSuffix = hash.toString(36).slice(-6) || 'x'
-    }
-  }
-
-  // Reserve space for suffix when calculating max base name length
-  const SUFFIX_LENGTH = serverIdSuffix ? serverIdSuffix.length + 1 : 0 // +1 for underscore
-  const MAX_BASE_LENGTH = 63 - SUFFIX_LENGTH
-
   // Combine server name and tool name
   let name = sanitizedTool
   if (!sanitizedTool.includes(sanitizedServer.slice(0, 7))) {
@@ -38,9 +20,9 @@ export function buildFunctionCallToolName(serverName: string, toolName: string,
   // Remove consecutive underscores/dashes (optional improvement)
   name = name.replace(/[_-]{2,}/g, '_')
 
-  // Truncate base name BEFORE adding suffix to ensure suffix is never cut off
-  if (name.length > MAX_BASE_LENGTH) {
-    name = name.slice(0, MAX_BASE_LENGTH)
+  // Truncate to 63 characters maximum
+  if (name.length > 63) {
+    name = name.slice(0, 63)
   }
 
   // Handle edge case: ensure we still have a valid name if truncation left invalid chars at edges
@@ -48,10 +30,5 @@ export function buildFunctionCallToolName(serverName: string, toolName: string,
     name = name.slice(0, -1)
   }
 
-  // Now append the suffix - it will always fit within 63 chars
-  if (serverIdSuffix) {
-    name = `${name}_${serverIdSuffix}`
-  }
-
   return name
 }

src/renderer/src/aiCore/chunk/AiSdkToChunkAdapter.ts

@@ -339,12 +339,14 @@ export class AiSdkToChunkAdapter {
           reasoning_content: final.reasoningContent || ''
         }
 
+        // Pass finishReason in BLOCK_COMPLETE for message-level tracking
         this.onChunk({
           type: ChunkType.BLOCK_COMPLETE,
           response: {
             ...baseResponse,
             usage: { ...usage },
-            metrics: metrics ? { ...metrics } : undefined
+            metrics: metrics ? { ...metrics } : undefined,
+            finishReason: chunk.finishReason
           }
         })
         this.onChunk({

src/renderer/src/aiCore/chunk/handleToolCallChunk.ts

@@ -212,9 +212,8 @@ export class ToolCallChunkHandler {
         description: toolName,
         type: 'builtin'
       } as BaseTool
-    } else if ((mcpTool = this.mcpTools.find((t) => t.id === toolName) as MCPTool)) {
+    } else if ((mcpTool = this.mcpTools.find((t) => t.name === toolName) as MCPTool)) {
       // 如果是客户端执行的 MCP 工具，沿用现有逻辑
-      // toolName is mcpTool.id (registered with id as key in convertMcpToolsToAiSdkTools)
       logger.info(`[ToolCallChunkHandler] Handling client-side MCP tool: ${toolName}`)
       // mcpTool = this.mcpTools.find((t) => t.name === toolName) as MCPTool
       // if (!mcpTool) {

src/renderer/src/aiCore/legacy/clients/BaseApiClient.ts

@@ -405,9 +405,6 @@ export abstract class BaseApiClient<
         if (!param.name?.trim()) {
           return acc
         }
-        // Parse JSON type parameters (Legacy API clients)
-        // Related: src/renderer/src/pages/settings/AssistantSettings/AssistantModelSettings.tsx:133-148
-        // The UI stores JSON type params as strings, this function parses them before sending to API
         if (param.type === 'json') {
           const value = param.value as string
           if (value === 'undefined') {

src/renderer/src/aiCore/legacy/clients/gemini/GeminiAPIClient.ts

@@ -46,7 +46,6 @@ import type {
   GeminiSdkRawOutput,
   GeminiSdkToolCall
 } from '@renderer/types/sdk'
-import { getTrailingApiVersion, withoutTrailingApiVersion } from '@renderer/utils'
 import { isToolUseModeFunction } from '@renderer/utils/assistant'
 import {
   geminiFunctionCallToMcpTool,
@@ -164,10 +163,6 @@ export class GeminiAPIClient extends BaseApiClient<
     return models
   }
 
-  override getBaseURL(): string {
-    return withoutTrailingApiVersion(super.getBaseURL())
-  }
-
   override async getSdkInstance() {
     if (this.sdkInstance) {
       return this.sdkInstance
@@ -193,13 +188,6 @@ export class GeminiAPIClient extends BaseApiClient<
     if (this.provider.isVertex) {
       return 'v1'
     }
-
-    // Extract trailing API version from the URL
-    const trailingVersion = getTrailingApiVersion(this.provider.apiHost || '')
-    if (trailingVersion) {
-      return trailingVersion
-    }
-
     return 'v1beta'
   }
 

src/renderer/src/aiCore/prepareParams/header.ts

@@ -7,7 +7,7 @@ import { isAwsBedrockProvider, isVertexProvider } from '@renderer/utils/provider
 // https://docs.claude.com/en/docs/build-with-claude/extended-thinking#interleaved-thinking
 const INTERLEAVED_THINKING_HEADER = 'interleaved-thinking-2025-05-14'
 // https://docs.claude.com/en/docs/build-with-claude/context-windows#1m-token-context-window
-// const CONTEXT_100M_HEADER = 'context-1m-2025-08-07'
+const CONTEXT_100M_HEADER = 'context-1m-2025-08-07'
 // https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/web-search
 const WEBSEARCH_HEADER = 'web-search-2025-03-05'
 
@@ -17,7 +17,7 @@ export function addAnthropicHeaders(assistant: Assistant, model: Model): string[
   if (
     isClaude45ReasoningModel(model) &&
     isToolUseModeFunction(assistant) &&
-    !(isVertexProvider(provider) || isAwsBedrockProvider(provider))
+    !(isVertexProvider(provider) && isAwsBedrockProvider(provider))
   ) {
     anthropicHeaders.push(INTERLEAVED_THINKING_HEADER)
   }
@@ -25,9 +25,7 @@ export function addAnthropicHeaders(assistant: Assistant, model: Model): string[
     if (isVertexProvider(provider) && assistant.enableWebSearch) {
       anthropicHeaders.push(WEBSEARCH_HEADER)
     }
-    // We may add it by user preference in assistant.settings instead of always adding it.
-    // See #11540, #11397
-    // anthropicHeaders.push(CONTEXT_100M_HEADER)
+    anthropicHeaders.push(CONTEXT_100M_HEADER)
   }
   return anthropicHeaders
 }

src/renderer/src/aiCore/prepareParams/parameterBuilder.ts

@@ -28,7 +28,6 @@ import { type Assistant, type MCPTool, type Provider } from '@renderer/types'
 import type { StreamTextParams } from '@renderer/types/aiCoreTypes'
 import { mapRegexToPatterns } from '@renderer/utils/blacklistMatchPattern'
 import { replacePromptVariables } from '@renderer/utils/prompt'
-import { isAwsBedrockProvider } from '@renderer/utils/provider'
 import type { ModelMessage, Tool } from 'ai'
 import { stepCountIs } from 'ai'
 
@@ -176,7 +175,7 @@ export async function buildStreamTextParams(
 
   let headers: Record<string, string | undefined> = options.requestOptions?.headers ?? {}
 
-  if (isAnthropicModel(model) && !isAwsBedrockProvider(provider)) {
+  if (isAnthropicModel(model)) {
     const newBetaHeaders = { 'anthropic-beta': addAnthropicHeaders(assistant, model).join(',') }
     headers = combineHeaders(headers, newBetaHeaders)
   }

src/renderer/src/aiCore/provider/__tests__/integratedRegistry.test.ts

@@ -1,4 +1,4 @@
-import type { Model, Provider } from '@renderer/types'
+import type { Provider } from '@renderer/types'
 import { describe, expect, it, vi } from 'vitest'
 
 import { getAiSdkProviderId } from '../factory'
@@ -68,18 +68,6 @@ function createTestProvider(id: string, type: string): Provider {
   } as Provider
 }
 
-function createAzureProvider(id: string, apiVersion?: string, model?: string): Provider {
-  return {
-    id,
-    type: 'azure-openai',
-    name: `Azure Test ${id}`,
-    apiKey: 'azure-test-key',
-    apiHost: 'azure-test-host',
-    apiVersion,
-    models: [{ id: model || 'gpt-4' } as Model]
-  }
-}
-
 describe('Integrated Provider Registry', () => {
   describe('Provider ID Resolution', () => {
     it('should resolve openrouter provider correctly', () => {
@@ -123,24 +111,6 @@ describe('Integrated Provider Registry', () => {
       const result = getAiSdkProviderId(unknownProvider)
       expect(result).toBe('unknown-provider')
     })
-
-    it('should handle Azure OpenAI providers correctly', () => {
-      const azureProvider = createAzureProvider('azure-test', '2024-02-15', 'gpt-4o')
-      const result = getAiSdkProviderId(azureProvider)
-      expect(result).toBe('azure')
-    })
-
-    it('should handle Azure OpenAI providers response endpoint correctly', () => {
-      const azureProvider = createAzureProvider('azure-test', 'v1', 'gpt-4o')
-      const result = getAiSdkProviderId(azureProvider)
-      expect(result).toBe('azure-responses')
-    })
-
-    it('should handle Azure provider Claude Models', () => {
-      const provider = createTestProvider('azure-anthropic', 'anthropic')
-      const result = getAiSdkProviderId(provider)
-      expect(result).toBe('azure-anthropic')
-    })
   })
 
   describe('Backward Compatibility', () => {

src/renderer/src/aiCore/trace/AiSdkSpanAdapter.ts

@@ -245,8 +245,8 @@ export class AiSdkSpanAdapter {
       'gen_ai.usage.output_tokens'
     ]
 
-    const promptTokens = attributes[inputsTokenKeys.find((key) => attributes[key]) || '']
-    const completionTokens = attributes[outputTokenKeys.find((key) => attributes[key]) || '']
+    const completionTokens = attributes[inputsTokenKeys.find((key) => attributes[key]) || '']
+    const promptTokens = attributes[outputTokenKeys.find((key) => attributes[key]) || '']
 
     if (completionTokens !== undefined || promptTokens !== undefined) {
       const usage: TokenUsage = {

src/renderer/src/aiCore/trace/__tests__/AiSdkSpanAdapter.test.ts

@@ -1,53 +0,0 @@
-import type { Span } from '@opentelemetry/api'
-import { SpanKind, SpanStatusCode } from '@opentelemetry/api'
-import { describe, expect, it, vi } from 'vitest'
-
-import { AiSdkSpanAdapter } from '../AiSdkSpanAdapter'
-
-vi.mock('@logger', () => ({
-  loggerService: {
-    withContext: () => ({
-      debug: vi.fn(),
-      error: vi.fn(),
-      info: vi.fn(),
-      warn: vi.fn()
-    })
-  }
-}))
-
-describe('AiSdkSpanAdapter', () => {
-  const createMockSpan = (attributes: Record<string, unknown>): Span => {
-    const span = {
-      spanContext: () => ({
-        traceId: 'trace-id',
-        spanId: 'span-id'
-      }),
-      _attributes: attributes,
-      _events: [],
-      name: 'test span',
-      status: { code: SpanStatusCode.OK },
-      kind: SpanKind.CLIENT,
-      startTime: [0, 0] as [number, number],
-      endTime: [0, 1] as [number, number],
-      ended: true,
-      parentSpanId: '',
-      links: []
-    }
-    return span as unknown as Span
-  }
-
-  it('maps prompt and completion usage tokens to the correct fields', () => {
-    const attributes = {
-      'ai.usage.promptTokens': 321,
-      'ai.usage.completionTokens': 654
-    }
-
-    const span = createMockSpan(attributes)
-    const result = AiSdkSpanAdapter.convertToSpanEntity({ span })
-
-    expect(result.usage).toBeDefined()
-    expect(result.usage?.prompt_tokens).toBe(321)
-    expect(result.usage?.completion_tokens).toBe(654)
-    expect(result.usage?.total_tokens).toBe(975)
-  })
-})

src/renderer/src/aiCore/utils/__tests__/mcp.test.ts

@@ -71,11 +71,10 @@ describe('mcp utils', () => {
       const result = setupToolsConfig(mcpTools)
 
       expect(result).not.toBeUndefined()
-      // Tools are now keyed by id (which includes serverId suffix) for uniqueness
-      expect(Object.keys(result!)).toEqual(['test-tool-1'])
-      expect(result!['test-tool-1']).toHaveProperty('description')
-      expect(result!['test-tool-1']).toHaveProperty('inputSchema')
-      expect(result!['test-tool-1']).toHaveProperty('execute')
+      expect(Object.keys(result!)).toEqual(['test-tool'])
+      expect(result!['test-tool']).toHaveProperty('description')
+      expect(result!['test-tool']).toHaveProperty('inputSchema')
+      expect(result!['test-tool']).toHaveProperty('execute')
     })
 
     it('should handle multiple MCP tools', () => {
@@ -110,8 +109,7 @@ describe('mcp utils', () => {
 
       expect(result).not.toBeUndefined()
       expect(Object.keys(result!)).toHaveLength(2)
-      // Tools are keyed by id for uniqueness
-      expect(Object.keys(result!)).toEqual(['tool1-id', 'tool2-id'])
+      expect(Object.keys(result!)).toEqual(['tool1', 'tool2'])
     })
   })
 
@@ -137,10 +135,9 @@ describe('mcp utils', () => {
 
       const result = convertMcpToolsToAiSdkTools(mcpTools)
 
-      // Tools are keyed by id for uniqueness when multiple server instances exist
-      expect(Object.keys(result)).toEqual(['get-weather-id'])
+      expect(Object.keys(result)).toEqual(['get-weather'])
 
-      const tool = result['get-weather-id'] as Tool
+      const tool = result['get-weather'] as Tool
       expect(tool.description).toBe('Get weather information')
       expect(tool.inputSchema).toBeDefined()
       expect(typeof tool.execute).toBe('function')
@@ -163,8 +160,8 @@ describe('mcp utils', () => {
 
       const result = convertMcpToolsToAiSdkTools(mcpTools)
 
-      expect(Object.keys(result)).toEqual(['no-desc-tool-id'])
-      const tool = result['no-desc-tool-id'] as Tool
+      expect(Object.keys(result)).toEqual(['no-desc-tool'])
+      const tool = result['no-desc-tool'] as Tool
       expect(tool.description).toBe('Tool from test-server')
     })
 
@@ -205,13 +202,13 @@ describe('mcp utils', () => {
 
       const result = convertMcpToolsToAiSdkTools(mcpTools)
 
-      expect(Object.keys(result)).toEqual(['complex-tool-id'])
-      const tool = result['complex-tool-id'] as Tool
+      expect(Object.keys(result)).toEqual(['complex-tool'])
+      const tool = result['complex-tool'] as Tool
       expect(tool.inputSchema).toBeDefined()
       expect(typeof tool.execute).toBe('function')
     })
 
-    it('should preserve tool id with special characters', () => {
+    it('should preserve tool names with special characters', () => {
       const mcpTools: MCPTool[] = [
         {
           id: 'special-tool-id',
@@ -228,8 +225,7 @@ describe('mcp utils', () => {
       ]
 
       const result = convertMcpToolsToAiSdkTools(mcpTools)
-      // Tools are keyed by id for uniqueness
-      expect(Object.keys(result)).toEqual(['special-tool-id'])
+      expect(Object.keys(result)).toEqual(['tool_with-special.chars'])
     })
 
     it('should handle multiple tools with different schemas', () => {
@@ -280,11 +276,10 @@ describe('mcp utils', () => {
 
       const result = convertMcpToolsToAiSdkTools(mcpTools)
 
-      // Tools are keyed by id for uniqueness
-      expect(Object.keys(result).sort()).toEqual(['boolean-tool-id', 'number-tool-id', 'string-tool-id'])
-      expect(result['string-tool-id']).toBeDefined()
-      expect(result['number-tool-id']).toBeDefined()
-      expect(result['boolean-tool-id']).toBeDefined()
+      expect(Object.keys(result).sort()).toEqual(['boolean-tool', 'number-tool', 'string-tool'])
+      expect(result['string-tool']).toBeDefined()
+      expect(result['number-tool']).toBeDefined()
+      expect(result['boolean-tool']).toBeDefined()
     })
   })
 
@@ -315,7 +310,7 @@ describe('mcp utils', () => {
       ]
 
       const tools = convertMcpToolsToAiSdkTools(mcpTools)
-      const tool = tools['test-exec-tool-id'] as Tool
+      const tool = tools['test-exec-tool'] as Tool
       const result = await tool.execute!({}, { messages: [], abortSignal: undefined, toolCallId: 'test-call-123' })
 
       expect(requestToolConfirmation).toHaveBeenCalled()
@@ -348,7 +343,7 @@ describe('mcp utils', () => {
       ]
 
       const tools = convertMcpToolsToAiSdkTools(mcpTools)
-      const tool = tools['cancelled-tool-id'] as Tool
+      const tool = tools['cancelled-tool'] as Tool
       const result = await tool.execute!({}, { messages: [], abortSignal: undefined, toolCallId: 'cancel-call-123' })
 
       expect(requestToolConfirmation).toHaveBeenCalled()
@@ -390,7 +385,7 @@ describe('mcp utils', () => {
       ]
 
       const tools = convertMcpToolsToAiSdkTools(mcpTools)
-      const tool = tools['error-tool-id'] as Tool
+      const tool = tools['error-tool'] as Tool
 
       await expect(
         tool.execute!({}, { messages: [], abortSignal: undefined, toolCallId: 'error-call-123' })
@@ -426,7 +421,7 @@ describe('mcp utils', () => {
       ]
 
       const tools = convertMcpToolsToAiSdkTools(mcpTools)
-      const tool = tools['auto-approve-tool-id'] as Tool
+      const tool = tools['auto-approve-tool'] as Tool
       const result = await tool.execute!({}, { messages: [], abortSignal: undefined, toolCallId: 'auto-call-123' })
 
       expect(requestToolConfirmation).not.toHaveBeenCalled()

src/renderer/src/aiCore/utils/__tests__/options.test.ts

@@ -154,10 +154,6 @@ vi.mock('../websearch', () => ({
   getWebSearchParams: vi.fn(() => ({ enable_search: true }))
 }))
 
-vi.mock('../../prepareParams/header', () => ({
-  addAnthropicHeaders: vi.fn(() => ['context-1m-2025-08-07'])
-}))
-
 const ensureWindowApi = () => {
   const globalWindow = window as any
   globalWindow.api = globalWindow.api || {}
@@ -637,64 +633,5 @@ describe('options utils', () => {
         expect(result.providerOptions).toHaveProperty('anthropic')
       })
     })
-
-    describe('AWS Bedrock provider', () => {
-      const bedrockProvider = {
-        id: 'bedrock',
-        name: 'AWS Bedrock',
-        type: 'aws-bedrock',
-        apiKey: 'test-key',
-        apiHost: 'https://bedrock.us-east-1.amazonaws.com',
-        models: [] as Model[]
-      } as Provider
-
-      const bedrockModel: Model = {
-        id: 'anthropic.claude-sonnet-4-20250514-v1:0',
-        name: 'Claude Sonnet 4',
-        provider: 'bedrock'
-      } as Model
-
-      it('should build basic Bedrock options', () => {
-        const result = buildProviderOptions(mockAssistant, bedrockModel, bedrockProvider, {
-          enableReasoning: false,
-          enableWebSearch: false,
-          enableGenerateImage: false
-        })
-
-        expect(result.providerOptions).toHaveProperty('bedrock')
-        expect(result.providerOptions.bedrock).toBeDefined()
-      })
-
-      it('should include anthropicBeta when Anthropic headers are needed', async () => {
-        const { addAnthropicHeaders } = await import('../../prepareParams/header')
-        vi.mocked(addAnthropicHeaders).mockReturnValue(['interleaved-thinking-2025-05-14', 'context-1m-2025-08-07'])
-
-        const result = buildProviderOptions(mockAssistant, bedrockModel, bedrockProvider, {
-          enableReasoning: false,
-          enableWebSearch: false,
-          enableGenerateImage: false
-        })
-
-        expect(result.providerOptions.bedrock).toHaveProperty('anthropicBeta')
-        expect(result.providerOptions.bedrock.anthropicBeta).toEqual([
-          'interleaved-thinking-2025-05-14',
-          'context-1m-2025-08-07'
-        ])
-      })
-
-      it('should include reasoning parameters when enabled', () => {
-        const result = buildProviderOptions(mockAssistant, bedrockModel, bedrockProvider, {
-          enableReasoning: true,
-          enableWebSearch: false,
-          enableGenerateImage: false
-        })
-
-        expect(result.providerOptions.bedrock).toHaveProperty('reasoningConfig')
-        expect(result.providerOptions.bedrock.reasoningConfig).toEqual({
-          type: 'enabled',
-          budgetTokens: 5000
-        })
-      })
-    })
   })
 })

src/renderer/src/aiCore/utils/__tests__/reasoning.test.ts

@@ -144,7 +144,7 @@ describe('reasoning utils', () => {
       expect(result).toEqual({})
     })
 
-    it('should not override reasoning for OpenRouter when reasoning effort undefined', async () => {
+    it('should disable reasoning for OpenRouter when no reasoning effort set', async () => {
       const { isReasoningModel } = await import('@renderer/config/models')
 
       vi.mocked(isReasoningModel).mockReturnValue(true)
@@ -161,29 +161,6 @@ describe('reasoning utils', () => {
         settings: {}
       } as Assistant
 
-      const result = getReasoningEffort(assistant, model)
-      expect(result).toEqual({})
-    })
-
-    it('should disable reasoning for OpenRouter when reasoning effort explicitly none', async () => {
-      const { isReasoningModel } = await import('@renderer/config/models')
-
-      vi.mocked(isReasoningModel).mockReturnValue(true)
-
-      const model: Model = {
-        id: 'anthropic/claude-sonnet-4',
-        name: 'Claude Sonnet 4',
-        provider: SystemProviderIds.openrouter
-      } as Model
-
-      const assistant: Assistant = {
-        id: 'test',
-        name: 'Test',
-        settings: {
-          reasoning_effort: 'none'
-        }
-      } as Assistant
-
       const result = getReasoningEffort(assistant, model)
       expect(result).toEqual({ reasoning: { enabled: false, exclude: true } })
     })
@@ -292,9 +269,7 @@ describe('reasoning utils', () => {
       const assistant: Assistant = {
         id: 'test',
         name: 'Test',
-        settings: {
-          reasoning_effort: 'none'
-        }
+        settings: {}
       } as Assistant
 
       const result = getReasoningEffort(assistant, model)

src/renderer/src/aiCore/utils/mcp.ts

@@ -28,9 +28,7 @@ export function convertMcpToolsToAiSdkTools(mcpTools: MCPTool[]): ToolSet {
   const tools: ToolSet = {}
 
   for (const mcpTool of mcpTools) {
-    // Use mcpTool.id (which includes serverId suffix) to ensure uniqueness
-    // when multiple instances of the same MCP server type are configured
-    tools[mcpTool.id] = tool({
+    tools[mcpTool.name] = tool({
       description: mcpTool.description || `Tool from ${mcpTool.serverName}`,
       inputSchema: jsonSchema(mcpTool.inputSchema as JSONSchema7),
       execute: async (params, { toolCallId }) => {

src/renderer/src/aiCore/utils/options.ts

@@ -36,7 +36,6 @@ import { isSupportServiceTierProvider, isSupportVerbosityProvider } from '@rende
 import type { JSONValue } from 'ai'
 import { t } from 'i18next'
 
-import { addAnthropicHeaders } from '../prepareParams/header'
 import { getAiSdkProviderId } from '../provider/factory'
 import { buildGeminiGenerateImageParams } from './image'
 import {
@@ -470,11 +469,6 @@ function buildBedrockProviderOptions(
     }
   }
 
-  const betaHeaders = addAnthropicHeaders(assistant, model)
-  if (betaHeaders.length > 0) {
-    providerOptions.anthropicBeta = betaHeaders
-  }
-
   return providerOptions
 }
 

src/renderer/src/aiCore/utils/reasoning.ts

@@ -16,8 +16,10 @@ import {
   isGPT5SeriesModel,
   isGPT51SeriesModel,
   isGrok4FastReasoningModel,
+  isGrokReasoningModel,
   isOpenAIDeepResearchModel,
   isOpenAIModel,
+  isOpenAIReasoningModel,
   isQwenAlwaysThinkModel,
   isQwenReasoningModel,
   isReasoningModel,
@@ -62,22 +64,30 @@ export function getReasoningEffort(assistant: Assistant, model: Model): Reasonin
   }
   const reasoningEffort = assistant?.settings?.reasoning_effort
 
-  // reasoningEffort is not set, no extra reasoning setting
-  // Generally, for every model which supports reasoning control, the reasoning effort won't be undefined.
-  // It's for some reasoning models that don't support reasoning control, such as deepseek reasoner.
-  if (!reasoningEffort) {
-    return {}
-  }
-
-  // Handle 'none' reasoningEffort. It's explicitly off.
-  if (reasoningEffort === 'none') {
+  // Handle undefined and 'none' reasoningEffort.
+  // TODO: They should be separated.
+  if (!reasoningEffort || reasoningEffort === 'none') {
     // openrouter: use reasoning
     if (model.provider === SystemProviderIds.openrouter) {
+      // Don't disable reasoning for Gemini models that support thinking tokens
+      if (isSupportedThinkingTokenGeminiModel(model) && !GEMINI_FLASH_MODEL_REGEX.test(model.id)) {
+        return {}
+      }
       // 'none' is not an available value for effort for now.
       // I think they should resolve this issue soon, so I'll just go ahead and use this value.
       if (isGPT51SeriesModel(model) && reasoningEffort === 'none') {
         return { reasoning: { effort: 'none' } }
       }
+      // Don't disable reasoning for models that require it
+      if (
+        isGrokReasoningModel(model) ||
+        isOpenAIReasoningModel(model) ||
+        isQwenAlwaysThinkModel(model) ||
+        model.id.includes('seed-oss') ||
+        model.id.includes('minimax-m2')
+      ) {
+        return {}
+      }
       return { reasoning: { enabled: false, exclude: true } }
     }
 
@@ -91,6 +101,11 @@ export function getReasoningEffort(assistant: Assistant, model: Model): Reasonin
       return { enable_thinking: false }
     }
 
+    // claude
+    if (isSupportedThinkingTokenClaudeModel(model)) {
+      return {}
+    }
+
     // gemini
     if (isSupportedThinkingTokenGeminiModel(model)) {
       if (GEMINI_FLASH_MODEL_REGEX.test(model.id)) {
@@ -103,10 +118,8 @@ export function getReasoningEffort(assistant: Assistant, model: Model): Reasonin
             }
           }
         }
-      } else {
-        logger.warn(`Model ${model.id} cannot disable reasoning. Fallback to empty reasoning param.`)
-        return {}
       }
+      return {}
     }
 
     // use thinking, doubao, zhipu, etc.
@@ -126,7 +139,6 @@ export function getReasoningEffort(assistant: Assistant, model: Model): Reasonin
       }
     }
 
-    logger.warn(`Model ${model.id} doesn't match any disable reasoning behavior. Fallback to empty reasoning param.`)
     return {}
   }
 
@@ -281,7 +293,6 @@ export function getReasoningEffort(assistant: Assistant, model: Model): Reasonin
   }
 
   // OpenRouter models, use reasoning
-  // FIXME: duplicated openrouter handling. remove one
   if (model.provider === SystemProviderIds.openrouter) {
     if (isSupportedReasoningEffortModel(model) || isSupportedThinkingTokenModel(model)) {
       return {
@@ -673,10 +684,6 @@ export function getCustomParameters(assistant: Assistant): Record<string, any> {
       if (!param.name?.trim()) {
         return acc
       }
-      // Parse JSON type parameters
-      // Related: src/renderer/src/pages/settings/AssistantSettings/AssistantModelSettings.tsx:133-148
-      // The UI stores JSON type params as strings (e.g., '{"key":"value"}')
-      // This function parses them into objects before sending to the API
       if (param.type === 'json') {
         const value = param.value as string
         if (value === 'undefined') {

src/renderer/src/assets/styles/ant.css

@@ -215,10 +215,6 @@
   border-top: none !important;
 }
 
-.ant-collapse-header-text {
-  overflow-x: hidden;
-}
-
 .ant-slider .ant-slider-handle::after {
   box-shadow: 0 1px 4px 0px rgb(128 128 128 / 50%) !important;
 }

src/renderer/src/components/ImageViewer.tsx

@@ -10,7 +10,6 @@ import {
 } from '@ant-design/icons'
 import { loggerService } from '@logger'
 import { download } from '@renderer/utils/download'
-import { convertImageToPng } from '@renderer/utils/image'
 import type { ImageProps as AntImageProps } from 'antd'
 import { Dropdown, Image as AntImage, Space } from 'antd'
 import { Base64 } from 'js-base64'
@@ -34,38 +33,39 @@ const ImageViewer: React.FC<ImageViewerProps> = ({ src, style, ...props }) => {
   // 复制图片到剪贴板
   const handleCopyImage = async (src: string) => {
     try {
-      let blob: Blob
-
       if (src.startsWith('data:')) {
         // 处理 base64 格式的图片
         const match = src.match(/^data:(image\/\w+);base64,(.+)$/)
         if (!match) throw new Error('Invalid base64 image format')
         const mimeType = match[1]
         const byteArray = Base64.toUint8Array(match[2])
-        blob = new Blob([byteArray], { type: mimeType })
+        const blob = new Blob([byteArray], { type: mimeType })
+        await navigator.clipboard.write([new ClipboardItem({ [mimeType]: blob })])
       } else if (src.startsWith('file://')) {
         // 处理本地文件路径
         const bytes = await window.api.fs.read(src)
         const mimeType = mime.getType(src) || 'application/octet-stream'
-        blob = new Blob([bytes], { type: mimeType })
+        const blob = new Blob([bytes], { type: mimeType })
+        await navigator.clipboard.write([
+          new ClipboardItem({
+            [mimeType]: blob
+          })
+        ])
       } else {
         // 处理 URL 格式的图片
         const response = await fetch(src)
-        blob = await response.blob()
+        const blob = await response.blob()
+
+        await navigator.clipboard.write([
+          new ClipboardItem({
+            [blob.type]: blob
+          })
+        ])
       }
 
-      // 统一转换为 PNG 以确保兼容性（剪贴板 API 不支持 JPEG）
-      const pngBlob = await convertImageToPng(blob)
-
-      const item = new ClipboardItem({
-        'image/png': pngBlob
-      })
-      await navigator.clipboard.write([item])
-
       window.toast.success(t('message.copy.success'))
     } catch (error) {
-      const err = error as Error
-      logger.error(`Failed to copy image: ${err.message}`, { stack: err.stack })
+      logger.error('Failed to copy image:', error as Error)
       window.toast.error(t('message.copy.failed'))
     }
   }

src/renderer/src/components/Popups/SelectModelPopup/api-model-popup.tsx

@@ -57,7 +57,7 @@ const PopupContainer: React.FC<Props> = ({ model, apiFilter, modelFilter, showTa
   const [_searchText, setSearchText] = useState('')
   const searchText = useDeferredValue(_searchText)
   const { models, isLoading } = useApiModels(apiFilter)
-  const adaptedModels = useMemo(() => models.map((model) => apiModelAdapter(model)), [models])
+  const adaptedModels = models.map((model) => apiModelAdapter(model))
 
   // 当前选中的模型ID
   const currentModelId = model ? model.id : ''

src/renderer/src/config/models/__tests__/reasoning.test.ts

@@ -309,14 +309,11 @@ describe('Ling Models', () => {
 describe('Claude & regional providers', () => {
   it('identifies claude 4.5 variants', () => {
     expect(isClaude45ReasoningModel(createModel({ id: 'claude-sonnet-4.5-preview' }))).toBe(true)
-    expect(isClaude4SeriesModel(createModel({ id: 'claude-sonnet-4-5@20250929' }))).toBe(true)
     expect(isClaude45ReasoningModel(createModel({ id: 'claude-3-sonnet' }))).toBe(false)
   })
 
   it('identifies claude 4 variants', () => {
     expect(isClaude4SeriesModel(createModel({ id: 'claude-opus-4' }))).toBe(true)
-    expect(isClaude4SeriesModel(createModel({ id: 'claude-sonnet-4@20250514' }))).toBe(true)
-    expect(isClaude4SeriesModel(createModel({ id: 'anthropic.claude-sonnet-4-20250514-v1:0' }))).toBe(true)
     expect(isClaude4SeriesModel(createModel({ id: 'claude-4.2-sonnet-variant' }))).toBe(false)
     expect(isClaude4SeriesModel(createModel({ id: 'claude-3-haiku' }))).toBe(false)
   })

src/renderer/src/config/models/__tests__/utils.test.ts

@@ -125,371 +125,195 @@ describe('model utils', () => {
     openAIWebSearchOnlyMock.mockReturnValue(false)
   })
 
-  describe('OpenAI model detection', () => {
-    describe('isOpenAILLMModel', () => {
-      it('returns false for undefined model', () => {
-        expect(isOpenAILLMModel(undefined as unknown as Model)).toBe(false)
-      })
+  it('detects OpenAI LLM models through reasoning and GPT prefix', () => {
+    expect(isOpenAILLMModel(undefined as unknown as Model)).toBe(false)
+    expect(isOpenAILLMModel(createModel({ id: 'gpt-4o-image' }))).toBe(false)
 
-      it('returns false for image generation models', () => {
-        expect(isOpenAILLMModel(createModel({ id: 'gpt-4o-image' }))).toBe(false)
-      })
+    reasoningMock.mockReturnValueOnce(true)
+    expect(isOpenAILLMModel(createModel({ id: 'o1-preview' }))).toBe(true)
 
-      it('returns true for reasoning models', () => {
-        reasoningMock.mockReturnValueOnce(true)
-        expect(isOpenAILLMModel(createModel({ id: 'o1-preview' }))).toBe(true)
-      })
+    expect(isOpenAILLMModel(createModel({ id: 'GPT-5-turbo' }))).toBe(true)
+  })
 
-      it('returns true for GPT-prefixed models', () => {
-        expect(isOpenAILLMModel(createModel({ id: 'GPT-5-turbo' }))).toBe(true)
-      })
+  it('detects OpenAI models via GPT prefix or reasoning support', () => {
+    expect(isOpenAIModel(createModel({ id: 'gpt-4.1' }))).toBe(true)
+    reasoningMock.mockReturnValueOnce(true)
+    expect(isOpenAIModel(createModel({ id: 'o3' }))).toBe(true)
+  })
+
+  it('evaluates support for flex service tier and alias helper', () => {
+    expect(isSupportFlexServiceTierModel(createModel({ id: 'o3' }))).toBe(true)
+    expect(isSupportFlexServiceTierModel(createModel({ id: 'o3-mini' }))).toBe(false)
+    expect(isSupportFlexServiceTierModel(createModel({ id: 'o4-mini' }))).toBe(true)
+    expect(isSupportFlexServiceTierModel(createModel({ id: 'gpt-5-preview' }))).toBe(true)
+    expect(isSupportedFlexServiceTier(createModel({ id: 'gpt-4o' }))).toBe(false)
+  })
+
+  it('detects verbosity support for GPT-5+ families', () => {
+    expect(isSupportVerbosityModel(createModel({ id: 'gpt-5' }))).toBe(true)
+    expect(isSupportVerbosityModel(createModel({ id: 'gpt-5-chat' }))).toBe(false)
+    expect(isSupportVerbosityModel(createModel({ id: 'gpt-5.1-preview' }))).toBe(true)
+  })
+
+  it('limits verbosity controls for GPT-5 Pro models', () => {
+    const proModel = createModel({ id: 'gpt-5-pro' })
+    const previewModel = createModel({ id: 'gpt-5-preview' })
+    expect(getModelSupportedVerbosity(proModel)).toEqual([undefined, 'high'])
+    expect(getModelSupportedVerbosity(previewModel)).toEqual([undefined, 'low', 'medium', 'high'])
+    expect(isGPT5ProModel(proModel)).toBe(true)
+    expect(isGPT5ProModel(previewModel)).toBe(false)
+  })
+
+  it('identifies OpenAI chat-completion-only models', () => {
+    expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'gpt-4o-search-preview' }))).toBe(true)
+    expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'o1-mini' }))).toBe(true)
+    expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'gpt-4o' }))).toBe(false)
+  })
+
+  it('filters unsupported OpenAI catalog entries', () => {
+    expect(isSupportedModel({ id: 'gpt-4', object: 'model' } as any)).toBe(true)
+    expect(isSupportedModel({ id: 'tts-1', object: 'model' } as any)).toBe(false)
+  })
+
+  it('calculates temperature/top-p support correctly', () => {
+    const model = createModel({ id: 'o1' })
+    reasoningMock.mockReturnValue(true)
+    expect(isNotSupportTemperatureAndTopP(model)).toBe(true)
+
+    const openWeight = createModel({ id: 'gpt-oss-debug' })
+    expect(isNotSupportTemperatureAndTopP(openWeight)).toBe(false)
+
+    const chatOnly = createModel({ id: 'o1-preview' })
+    reasoningMock.mockReturnValue(false)
+    expect(isNotSupportTemperatureAndTopP(chatOnly)).toBe(true)
+
+    const qwenMt = createModel({ id: 'qwen-mt-large', provider: 'aliyun' })
+    expect(isNotSupportTemperatureAndTopP(qwenMt)).toBe(true)
+  })
+
+  it('handles gemma and gemini detections plus zhipu tagging', () => {
+    expect(isGemmaModel(createModel({ id: 'Gemma-3-27B' }))).toBe(true)
+    expect(isGemmaModel(createModel({ group: 'Gemma' }))).toBe(true)
+    expect(isGemmaModel(createModel({ id: 'gpt-4o' }))).toBe(false)
+
+    expect(isGeminiModel(createModel({ id: 'Gemini-2.0' }))).toBe(true)
+
+    expect(isZhipuModel(createModel({ provider: 'zhipu' }))).toBe(true)
+    expect(isZhipuModel(createModel({ provider: 'openai' }))).toBe(false)
+  })
+
+  it('groups qwen models by prefix', () => {
+    const qwen = createModel({ id: 'Qwen-7B', provider: 'qwen', name: 'Qwen-7B' })
+    const qwenOmni = createModel({ id: 'qwen2.5-omni', name: 'qwen2.5-omni' })
+    const other = createModel({ id: 'deepseek-v3', group: 'DeepSeek' })
+
+    const grouped = groupQwenModels([qwen, qwenOmni, other])
+    expect(Object.keys(grouped)).toContain('qwen-7b')
+    expect(Object.keys(grouped)).toContain('qwen2.5')
+    expect(grouped.DeepSeek).toContain(other)
+  })
+
+  it('aggregates boolean helpers based on regex rules', () => {
+    expect(isAnthropicModel(createModel({ id: 'claude-3.5' }))).toBe(true)
+    expect(isQwenMTModel(createModel({ id: 'qwen-mt-plus' }))).toBe(true)
+    expect(isNotSupportSystemMessageModel(createModel({ id: 'gemma-moe' }))).toBe(true)
+    expect(isOpenAIOpenWeightModel(createModel({ id: 'gpt-oss-free' }))).toBe(true)
+  })
+
+  describe('isNotSupportedTextDelta', () => {
+    it('returns true for qwen-mt-turbo and qwen-mt-plus models', () => {
+      // qwen-mt series that don't support text delta
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-turbo' }))).toBe(true)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-plus' }))).toBe(true)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'Qwen-MT-Turbo' }))).toBe(true)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'QWEN-MT-PLUS' }))).toBe(true)
     })
 
-    describe('isOpenAIModel', () => {
-      it('detects models via GPT prefix', () => {
-        expect(isOpenAIModel(createModel({ id: 'gpt-4.1' }))).toBe(true)
-      })
+    it('returns false for qwen-mt-flash and other models', () => {
+      // qwen-mt-flash supports text delta
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-flash' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'Qwen-MT-Flash' }))).toBe(false)
 
-      it('detects models via reasoning support', () => {
-        reasoningMock.mockReturnValueOnce(true)
-        expect(isOpenAIModel(createModel({ id: 'o3' }))).toBe(true)
-      })
+      // Legacy qwen models without mt prefix (support text delta)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-turbo' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-plus' }))).toBe(false)
+
+      // Other qwen models
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-max' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen2.5-72b' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-vl-plus' }))).toBe(false)
+
+      // Non-qwen models
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'gpt-4o' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'claude-3.5' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'glm-4-plus' }))).toBe(false)
     })
 
-    describe('isOpenAIChatCompletionOnlyModel', () => {
-      it('identifies chat-completion-only models', () => {
-        expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'gpt-4o-search-preview' }))).toBe(true)
-        expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'o1-mini' }))).toBe(true)
-      })
+    it('handles models with version suffixes', () => {
+      // qwen-mt models with version suffixes
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-turbo-1201' }))).toBe(true)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-plus-0828' }))).toBe(true)
 
-      it('returns false for general models', () => {
-        expect(isOpenAIChatCompletionOnlyModel(createModel({ id: 'gpt-4o' }))).toBe(false)
-      })
+      // Legacy qwen models with version suffixes (support text delta)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-turbo-0828' }))).toBe(false)
+      expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-plus-latest' }))).toBe(false)
     })
   })
 
-  describe('GPT-5 family detection', () => {
-    describe('isGPT5SeriesModel', () => {
-      it('returns true for GPT-5 models', () => {
-        expect(isGPT5SeriesModel(createModel({ id: 'gpt-5-preview' }))).toBe(true)
-      })
-
-      it('returns false for GPT-5.1 models', () => {
-        expect(isGPT5SeriesModel(createModel({ id: 'gpt-5.1-preview' }))).toBe(false)
-      })
-    })
-
-    describe('isGPT51SeriesModel', () => {
-      it('returns true for GPT-5.1 models', () => {
-        expect(isGPT51SeriesModel(createModel({ id: 'gpt-5.1-mini' }))).toBe(true)
-      })
-    })
-
-    describe('isGPT5SeriesReasoningModel', () => {
-      it('returns true for GPT-5 reasoning models', () => {
-        expect(isGPT5SeriesReasoningModel(createModel({ id: 'gpt-5' }))).toBe(true)
-      })
-      it('returns false for gpt-5-chat', () => {
-        expect(isGPT5SeriesReasoningModel(createModel({ id: 'gpt-5-chat' }))).toBe(false)
-      })
-    })
-
-    describe('isGPT5ProModel', () => {
-      it('returns true for GPT-5 Pro models', () => {
-        expect(isGPT5ProModel(createModel({ id: 'gpt-5-pro' }))).toBe(true)
-      })
-
-      it('returns false for non-Pro GPT-5 models', () => {
-        expect(isGPT5ProModel(createModel({ id: 'gpt-5-preview' }))).toBe(false)
-      })
-    })
+  it('evaluates GPT-5 family helpers', () => {
+    expect(isGPT5SeriesModel(createModel({ id: 'gpt-5-preview' }))).toBe(true)
+    expect(isGPT5SeriesModel(createModel({ id: 'gpt-5.1-preview' }))).toBe(false)
+    expect(isGPT51SeriesModel(createModel({ id: 'gpt-5.1-mini' }))).toBe(true)
+    expect(isGPT5SeriesReasoningModel(createModel({ id: 'gpt-5-prompt' }))).toBe(true)
+    expect(isSupportVerbosityModel(createModel({ id: 'gpt-5-chat' }))).toBe(false)
   })
 
-  describe('Verbosity support', () => {
-    describe('isSupportVerbosityModel', () => {
-      it('returns true for GPT-5 models', () => {
-        expect(isSupportVerbosityModel(createModel({ id: 'gpt-5' }))).toBe(true)
-      })
+  it('wraps generate/vision helpers that operate on arrays', () => {
+    const models = [createModel({ id: 'gpt-4o' }), createModel({ id: 'gpt-4o-mini' })]
+    expect(isVisionModels(models)).toBe(true)
+    visionMock.mockReturnValueOnce(true).mockReturnValueOnce(false)
+    expect(isVisionModels(models)).toBe(false)
 
-      it('returns false for GPT-5 chat models', () => {
-        expect(isSupportVerbosityModel(createModel({ id: 'gpt-5-chat' }))).toBe(false)
-      })
-
-      it('returns true for GPT-5.1 models', () => {
-        expect(isSupportVerbosityModel(createModel({ id: 'gpt-5.1-preview' }))).toBe(true)
-      })
-    })
-
-    describe('getModelSupportedVerbosity', () => {
-      it('returns only "high" for GPT-5 Pro models', () => {
-        expect(getModelSupportedVerbosity(createModel({ id: 'gpt-5-pro' }))).toEqual([undefined, 'high'])
-        expect(getModelSupportedVerbosity(createModel({ id: 'gpt-5-pro-2025-10-06' }))).toEqual([undefined, 'high'])
-      })
-
-      it('returns all levels for non-Pro GPT-5 models', () => {
-        const previewModel = createModel({ id: 'gpt-5-preview' })
-        expect(getModelSupportedVerbosity(previewModel)).toEqual([undefined, 'low', 'medium', 'high'])
-      })
-
-      it('returns all levels for GPT-5.1 models', () => {
-        const gpt51Model = createModel({ id: 'gpt-5.1-preview' })
-        expect(getModelSupportedVerbosity(gpt51Model)).toEqual([undefined, 'low', 'medium', 'high'])
-      })
-
-      it('returns only undefined for non-GPT-5 models', () => {
-        expect(getModelSupportedVerbosity(createModel({ id: 'gpt-4o' }))).toEqual([undefined])
-        expect(getModelSupportedVerbosity(createModel({ id: 'claude-3.5' }))).toEqual([undefined])
-      })
-
-      it('returns only undefined for undefiend/null input', () => {
-        expect(getModelSupportedVerbosity(undefined)).toEqual([undefined])
-        expect(getModelSupportedVerbosity(null)).toEqual([undefined])
-      })
-    })
+    expect(isGenerateImageModels(models)).toBe(true)
+    generateImageMock.mockReturnValueOnce(true).mockReturnValueOnce(false)
+    expect(isGenerateImageModels(models)).toBe(false)
   })
 
-  describe('Flex service tier support', () => {
-    describe('isSupportFlexServiceTierModel', () => {
-      it('returns true for supported models', () => {
-        expect(isSupportFlexServiceTierModel(createModel({ id: 'o3' }))).toBe(true)
-        expect(isSupportFlexServiceTierModel(createModel({ id: 'o4-mini' }))).toBe(true)
-        expect(isSupportFlexServiceTierModel(createModel({ id: 'gpt-5-preview' }))).toBe(true)
-      })
+  it('filters models for agent usage', () => {
+    expect(agentModelFilter(createModel())).toBe(true)
 
-      it('returns false for unsupported models', () => {
-        expect(isSupportFlexServiceTierModel(createModel({ id: 'o3-mini' }))).toBe(false)
-      })
-    })
+    embeddingMock.mockReturnValueOnce(true)
+    expect(agentModelFilter(createModel({ id: 'text-embedding' }))).toBe(false)
 
-    describe('isSupportedFlexServiceTier', () => {
-      it('returns false for non-flex models', () => {
-        expect(isSupportedFlexServiceTier(createModel({ id: 'gpt-4o' }))).toBe(false)
-      })
-    })
+    embeddingMock.mockReturnValue(false)
+    rerankMock.mockReturnValueOnce(true)
+    expect(agentModelFilter(createModel({ id: 'rerank' }))).toBe(false)
+
+    rerankMock.mockReturnValue(false)
+    textToImageMock.mockReturnValueOnce(true)
+    expect(agentModelFilter(createModel({ id: 'gpt-image-1' }))).toBe(false)
   })
 
-  describe('Temperature and top-p support', () => {
-    describe('isNotSupportTemperatureAndTopP', () => {
-      it('returns true for reasoning models', () => {
-        const model = createModel({ id: 'o1' })
-        reasoningMock.mockReturnValue(true)
-        expect(isNotSupportTemperatureAndTopP(model)).toBe(true)
-      })
+  it('identifies models with maximum temperature of 1.0', () => {
+    // Zhipu models should have max temperature of 1.0
+    expect(isMaxTemperatureOneModel(createModel({ id: 'glm-4' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'GLM-4-Plus' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'glm-3-turbo' }))).toBe(true)
 
-      it('returns false for open weight models', () => {
-        const openWeight = createModel({ id: 'gpt-oss-debug' })
-        expect(isNotSupportTemperatureAndTopP(openWeight)).toBe(false)
-      })
+    // Anthropic models should have max temperature of 1.0
+    expect(isMaxTemperatureOneModel(createModel({ id: 'claude-3.5-sonnet' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'Claude-3-opus' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'claude-2.1' }))).toBe(true)
 
-      it('returns true for chat-only models without reasoning', () => {
-        const chatOnly = createModel({ id: 'o1-preview' })
-        reasoningMock.mockReturnValue(false)
-        expect(isNotSupportTemperatureAndTopP(chatOnly)).toBe(true)
-      })
+    // Moonshot models should have max temperature of 1.0
+    expect(isMaxTemperatureOneModel(createModel({ id: 'moonshot-1.0' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'kimi-k2-thinking' }))).toBe(true)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'Moonshot-Pro' }))).toBe(true)
 
-      it('returns true for Qwen MT models', () => {
-        const qwenMt = createModel({ id: 'qwen-mt-large', provider: 'aliyun' })
-        expect(isNotSupportTemperatureAndTopP(qwenMt)).toBe(true)
-      })
-    })
-  })
-
-  describe('Text delta support', () => {
-    describe('isNotSupportTextDeltaModel', () => {
-      it('returns true for qwen-mt-turbo and qwen-mt-plus models', () => {
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-turbo' }))).toBe(true)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-plus' }))).toBe(true)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'Qwen-MT-Turbo' }))).toBe(true)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'QWEN-MT-PLUS' }))).toBe(true)
-      })
-
-      it('returns false for qwen-mt-flash and other models', () => {
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-flash' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'Qwen-MT-Flash' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-turbo' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-plus' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-max' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen2.5-72b' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-vl-plus' }))).toBe(false)
-      })
-
-      it('returns false for non-qwen models', () => {
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'gpt-4o' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'claude-3.5' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'glm-4-plus' }))).toBe(false)
-      })
-
-      it('handles models with version suffixes', () => {
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-turbo-1201' }))).toBe(true)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-mt-plus-0828' }))).toBe(true)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-turbo-0828' }))).toBe(false)
-        expect(isNotSupportTextDeltaModel(createModel({ id: 'qwen-plus-latest' }))).toBe(false)
-      })
-    })
-  })
-
-  describe('Model provider detection', () => {
-    describe('isGemmaModel', () => {
-      it('detects Gemma models by ID', () => {
-        expect(isGemmaModel(createModel({ id: 'Gemma-3-27B' }))).toBe(true)
-      })
-
-      it('detects Gemma models by group', () => {
-        expect(isGemmaModel(createModel({ group: 'Gemma' }))).toBe(true)
-      })
-
-      it('returns false for non-Gemma models', () => {
-        expect(isGemmaModel(createModel({ id: 'gpt-4o' }))).toBe(false)
-      })
-    })
-
-    describe('isGeminiModel', () => {
-      it('detects Gemini models', () => {
-        expect(isGeminiModel(createModel({ id: 'Gemini-2.0' }))).toBe(true)
-      })
-    })
-
-    describe('isZhipuModel', () => {
-      it('detects Zhipu models by provider', () => {
-        expect(isZhipuModel(createModel({ provider: 'zhipu' }))).toBe(true)
-      })
-
-      it('returns false for non-Zhipu models', () => {
-        expect(isZhipuModel(createModel({ provider: 'openai' }))).toBe(false)
-      })
-    })
-
-    describe('isAnthropicModel', () => {
-      it('detects Anthropic models', () => {
-        expect(isAnthropicModel(createModel({ id: 'claude-3.5' }))).toBe(true)
-      })
-    })
-
-    describe('isQwenMTModel', () => {
-      it('detects Qwen MT models', () => {
-        expect(isQwenMTModel(createModel({ id: 'qwen-mt-plus' }))).toBe(true)
-      })
-    })
-
-    describe('isOpenAIOpenWeightModel', () => {
-      it('detects OpenAI open weight models', () => {
-        expect(isOpenAIOpenWeightModel(createModel({ id: 'gpt-oss-free' }))).toBe(true)
-      })
-    })
-  })
-
-  describe('System message support', () => {
-    describe('isNotSupportSystemMessageModel', () => {
-      it('returns true for models that do not support system messages', () => {
-        expect(isNotSupportSystemMessageModel(createModel({ id: 'gemma-moe' }))).toBe(true)
-      })
-    })
-  })
-
-  describe('Model grouping', () => {
-    describe('groupQwenModels', () => {
-      it('groups qwen models by prefix', () => {
-        const qwen = createModel({ id: 'Qwen-7B', provider: 'qwen', name: 'Qwen-7B' })
-        const qwenOmni = createModel({ id: 'qwen2.5-omni', name: 'qwen2.5-omni' })
-        const other = createModel({ id: 'deepseek-v3', group: 'DeepSeek' })
-
-        const grouped = groupQwenModels([qwen, qwenOmni, other])
-        expect(Object.keys(grouped)).toContain('qwen-7b')
-        expect(Object.keys(grouped)).toContain('qwen2.5')
-        expect(grouped.DeepSeek).toContain(other)
-      })
-    })
-  })
-
-  describe('Vision and image generation', () => {
-    describe('isVisionModels', () => {
-      it('returns true when all models support vision', () => {
-        const models = [createModel({ id: 'gpt-4o' }), createModel({ id: 'gpt-4o-mini' })]
-        expect(isVisionModels(models)).toBe(true)
-      })
-
-      it('returns false when some models do not support vision', () => {
-        const models = [createModel({ id: 'gpt-4o' }), createModel({ id: 'gpt-4o-mini' })]
-        visionMock.mockReturnValueOnce(true).mockReturnValueOnce(false)
-        expect(isVisionModels(models)).toBe(false)
-      })
-    })
-
-    describe('isGenerateImageModels', () => {
-      it('returns true when all models support image generation', () => {
-        const models = [createModel({ id: 'gpt-4o' }), createModel({ id: 'gpt-4o-mini' })]
-        expect(isGenerateImageModels(models)).toBe(true)
-      })
-
-      it('returns false when some models do not support image generation', () => {
-        const models = [createModel({ id: 'gpt-4o' }), createModel({ id: 'gpt-4o-mini' })]
-        generateImageMock.mockReturnValueOnce(true).mockReturnValueOnce(false)
-        expect(isGenerateImageModels(models)).toBe(false)
-      })
-    })
-  })
-
-  describe('Model filtering', () => {
-    describe('isSupportedModel', () => {
-      it('filters supported OpenAI catalog entries', () => {
-        expect(isSupportedModel({ id: 'gpt-4', object: 'model' } as any)).toBe(true)
-      })
-
-      it('filters unsupported OpenAI catalog entries', () => {
-        expect(isSupportedModel({ id: 'tts-1', object: 'model' } as any)).toBe(false)
-      })
-    })
-
-    describe('agentModelFilter', () => {
-      it('returns true for regular models', () => {
-        expect(agentModelFilter(createModel())).toBe(true)
-      })
-
-      it('filters out embedding models', () => {
-        embeddingMock.mockReturnValueOnce(true)
-        expect(agentModelFilter(createModel({ id: 'text-embedding' }))).toBe(false)
-      })
-
-      it('filters out rerank models', () => {
-        embeddingMock.mockReturnValue(false)
-        rerankMock.mockReturnValueOnce(true)
-        expect(agentModelFilter(createModel({ id: 'rerank' }))).toBe(false)
-      })
-
-      it('filters out text-to-image models', () => {
-        rerankMock.mockReturnValue(false)
-        textToImageMock.mockReturnValueOnce(true)
-        expect(agentModelFilter(createModel({ id: 'gpt-image-1' }))).toBe(false)
-      })
-    })
-  })
-
-  describe('Temperature limits', () => {
-    describe('isMaxTemperatureOneModel', () => {
-      it('returns true for Zhipu models', () => {
-        expect(isMaxTemperatureOneModel(createModel({ id: 'glm-4' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'GLM-4-Plus' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'glm-3-turbo' }))).toBe(true)
-      })
-
-      it('returns true for Anthropic models', () => {
-        expect(isMaxTemperatureOneModel(createModel({ id: 'claude-3.5-sonnet' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'Claude-3-opus' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'claude-2.1' }))).toBe(true)
-      })
-
-      it('returns true for Moonshot models', () => {
-        expect(isMaxTemperatureOneModel(createModel({ id: 'moonshot-1.0' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'kimi-k2-thinking' }))).toBe(true)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'Moonshot-Pro' }))).toBe(true)
-      })
-
-      it('returns false for other models', () => {
-        expect(isMaxTemperatureOneModel(createModel({ id: 'gpt-4o' }))).toBe(false)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'gpt-4-turbo' }))).toBe(false)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'qwen-max' }))).toBe(false)
-        expect(isMaxTemperatureOneModel(createModel({ id: 'gemini-pro' }))).toBe(false)
-      })
-    })
+    // Other models should return false
+    expect(isMaxTemperatureOneModel(createModel({ id: 'gpt-4o' }))).toBe(false)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'gpt-4-turbo' }))).toBe(false)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'qwen-max' }))).toBe(false)
+    expect(isMaxTemperatureOneModel(createModel({ id: 'gemini-pro' }))).toBe(false)
   })
 })

src/renderer/src/config/models/reasoning.ts

@@ -396,11 +396,7 @@ export function isClaude45ReasoningModel(model: Model): boolean {
 
 export function isClaude4SeriesModel(model: Model): boolean {
   const modelId = getLowerBaseModelName(model.id, '/')
-  // Supports various formats including:
-  // - Direct API: claude-sonnet-4, claude-opus-4-20250514
-  // - GCP Vertex AI: claude-sonnet-4@20250514
-  // - AWS Bedrock: anthropic.claude-sonnet-4-20250514-v1:0
-  const regex = /claude-(sonnet|opus|haiku)-4(?:[.-]\d+)?(?:[@\-:][\w\-:]+)?$/i
+  const regex = /claude-(sonnet|opus|haiku)-4(?:[.-]\d+)?(?:-[\w-]+)?$/i
   return regex.test(modelId)
 }
 
@@ -460,19 +456,16 @@ export const isSupportedThinkingTokenZhipuModel = (model: Model): boolean => {
 }
 
 export const isDeepSeekHybridInferenceModel = (model: Model) => {
-  const { idResult, nameResult } = withModelIdAndNameAsId(model, (model) => {
-    const modelId = getLowerBaseModelName(model.id)
-    // deepseek官方使用chat和reasoner做推理控制，其他provider需要单独判断，id可能会有所差别
-    // openrouter: deepseek/deepseek-chat-v3.1 不知道会不会有其他provider仿照ds官方分出一个同id的作为非思考模式的模型，这里有风险
-    // Matches: "deepseek-v3" followed by ".digit" or "-digit".
-    // Optionally, this can be followed by ".alphanumeric_sequence" or "-alphanumeric_sequence"
-    // until the end of the string.
-    // Examples: deepseek-v3.1, deepseek-v3-1, deepseek-v3.1.2, deepseek-v3.1-alpha
-    // Does NOT match: deepseek-v3.123 (missing separator after '1'), deepseek-v3.x (x isn't a digit)
-    // TODO: move to utils and add test cases
-    return /deepseek-v3(?:\.\d|-\d)(?:(\.|-)\w+)?$/.test(modelId) || modelId.includes('deepseek-chat-v3.1')
-  })
-  return idResult || nameResult
+  const modelId = getLowerBaseModelName(model.id)
+  // deepseek官方使用chat和reasoner做推理控制，其他provider需要单独判断，id可能会有所差别
+  // openrouter: deepseek/deepseek-chat-v3.1 不知道会不会有其他provider仿照ds官方分出一个同id的作为非思考模式的模型，这里有风险
+  // Matches: "deepseek-v3" followed by ".digit" or "-digit".
+  // Optionally, this can be followed by ".alphanumeric_sequence" or "-alphanumeric_sequence"
+  // until the end of the string.
+  // Examples: deepseek-v3.1, deepseek-v3-1, deepseek-v3.1.2, deepseek-v3.1-alpha
+  // Does NOT match: deepseek-v3.123 (missing separator after '1'), deepseek-v3.x (x isn't a digit)
+  // TODO: move to utils and add test cases
+  return /deepseek-v3(?:\.\d|-\d)(?:(\.|-)\w+)?$/.test(modelId) || modelId.includes('deepseek-chat-v3.1')
 }
 
 export const isLingReasoningModel = (model?: Model): boolean => {
@@ -526,6 +519,7 @@ export function isReasoningModel(model?: Model): boolean {
       REASONING_REGEX.test(model.name) ||
       isSupportedThinkingTokenDoubaoModel(model) ||
       isDeepSeekHybridInferenceModel(model) ||
+      isDeepSeekHybridInferenceModel({ ...model, id: model.name }) ||
       false
     )
   }

src/renderer/src/config/models/utils.ts

@@ -4,14 +4,7 @@ import { type Model, SystemProviderIds } from '@renderer/types'
 import type { OpenAIVerbosity, ValidOpenAIVerbosity } from '@renderer/types/aiCoreTypes'
 import { getLowerBaseModelName } from '@renderer/utils'
 
-import {
-  isGPT5ProModel,
-  isGPT5SeriesModel,
-  isGPT51SeriesModel,
-  isOpenAIChatCompletionOnlyModel,
-  isOpenAIOpenWeightModel,
-  isOpenAIReasoningModel
-} from './openai'
+import { isOpenAIChatCompletionOnlyModel, isOpenAIOpenWeightModel, isOpenAIReasoningModel } from './openai'
 import { isQwenMTModel } from './qwen'
 import { isGenerateImageModel, isTextToImageModel, isVisionModel } from './vision'
 export const NOT_SUPPORTED_REGEX = /(?:^tts|whisper|speech)/i
@@ -130,46 +123,21 @@ export const isNotSupportSystemMessageModel = (model: Model): boolean => {
   return isQwenMTModel(model) || isGemmaModel(model)
 }
 
-// Verbosity settings is only supported by GPT-5 and newer models
-// Specifically, GPT-5 and GPT-5.1 for now
+// GPT-5 verbosity configuration
 // gpt-5-pro only supports 'high', other GPT-5 models support all levels
-const MODEL_SUPPORTED_VERBOSITY: readonly {
-  readonly validator: (model: Model) => boolean
-  readonly values: readonly ValidOpenAIVerbosity[]
-}[] = [
-  // gpt-5-pro
-  { validator: isGPT5ProModel, values: ['high'] },
-  // gpt-5 except gpt-5-pro
-  {
-    validator: (model: Model) => isGPT5SeriesModel(model) && !isGPT5ProModel(model),
-    values: ['low', 'medium', 'high']
-  },
-  // gpt-5.1
-  { validator: isGPT51SeriesModel, values: ['low', 'medium', 'high'] }
-]
+export const MODEL_SUPPORTED_VERBOSITY: Record<string, ValidOpenAIVerbosity[]> = {
+  'gpt-5-pro': ['high'],
+  default: ['low', 'medium', 'high']
+} as const
 
-/**
- * Returns the list of supported verbosity levels for the given model.
- * If the model is not recognized as a GPT-5 series model, only `undefined` is returned.
- * For GPT-5-pro, only 'high' is supported; for other GPT-5 models, 'low', 'medium', and 'high' are supported.
- * For GPT-5.1 series models, 'low', 'medium', and 'high' are supported.
- * @param model - The model to check
- * @returns An array of supported verbosity levels, always including `undefined` as the first element
- */
-export const getModelSupportedVerbosity = (model: Model | undefined | null): OpenAIVerbosity[] => {
-  if (!model) {
-    return [undefined]
+export const getModelSupportedVerbosity = (model: Model): OpenAIVerbosity[] => {
+  const modelId = getLowerBaseModelName(model.id)
+  let supportedValues: ValidOpenAIVerbosity[]
+  if (modelId.includes('gpt-5-pro')) {
+    supportedValues = MODEL_SUPPORTED_VERBOSITY['gpt-5-pro']
+  } else {
+    supportedValues = MODEL_SUPPORTED_VERBOSITY.default
   }
-
-  let supportedValues: ValidOpenAIVerbosity[] = []
-
-  for (const { validator, values } of MODEL_SUPPORTED_VERBOSITY) {
-    if (validator(model)) {
-      supportedValues = [...values]
-      break
-    }
-  }
-
   return [undefined, ...supportedValues]
 }
 

src/renderer/src/hooks/useScrollPosition.ts

@@ -1,5 +1,5 @@
 import { throttle } from 'lodash'
-import { useEffect, useMemo, useRef } from 'react'
+import { useEffect, useRef } from 'react'
 
 import { useTimer } from './useTimer'
 
@@ -12,18 +12,13 @@ import { useTimer } from './useTimer'
  */
 export default function useScrollPosition(key: string, throttleWait?: number) {
   const containerRef = useRef<HTMLDivElement>(null)
-  const scrollKey = useMemo(() => `scroll:${key}`, [key])
-  const scrollKeyRef = useRef(scrollKey)
+  const scrollKey = `scroll:${key}`
   const { setTimeoutTimer } = useTimer()
 
-  useEffect(() => {
-    scrollKeyRef.current = scrollKey
-  }, [scrollKey])
-
   const handleScroll = throttle(() => {
     const position = containerRef.current?.scrollTop ?? 0
     window.requestAnimationFrame(() => {
-      window.keyv.set(scrollKeyRef.current, position)
+      window.keyv.set(scrollKey, position)
     })
   }, throttleWait ?? 100)
 
@@ -33,9 +28,5 @@ export default function useScrollPosition(key: string, throttleWait?: number) {
     setTimeoutTimer('scrollEffect', scroll, 50)
   }, [scrollKey, setTimeoutTimer])
 
-  useEffect(() => {
-    return () => handleScroll.cancel()
-  }, [handleScroll])
-
   return { containerRef, handleScroll }
 }

src/renderer/src/hooks/useTimer.ts

@@ -1,4 +1,4 @@
-import { useCallback, useEffect, useRef } from 'react'
+import { useEffect, useRef } from 'react'
 
 /**
  * 定时器管理 Hook，用于管理 setTimeout 和 setInterval 定时器，支持通过 key 来标识不同的定时器
@@ -43,38 +43,10 @@ export const useTimer = () => {
   const timeoutMapRef = useRef(new Map<string, NodeJS.Timeout>())
   const intervalMapRef = useRef(new Map<string, NodeJS.Timeout>())
 
-  /**
-   * 清除指定 key 的 setTimeout 定时器
-   * @param key - 定时器标识符
-   */
-  const clearTimeoutTimer = useCallback((key: string) => {
-    clearTimeout(timeoutMapRef.current.get(key))
-    timeoutMapRef.current.delete(key)
-  }, [])
-
-  /**
-   * 清除指定 key 的 setInterval 定时器
-   * @param key - 定时器标识符
-   */
-  const clearIntervalTimer = useCallback((key: string) => {
-    clearInterval(intervalMapRef.current.get(key))
-    intervalMapRef.current.delete(key)
-  }, [])
-
-  /**
-   * 清除所有定时器，包括 setTimeout 和 setInterval
-   */
-  const clearAllTimers = useCallback(() => {
-    timeoutMapRef.current.forEach((timer) => clearTimeout(timer))
-    intervalMapRef.current.forEach((timer) => clearInterval(timer))
-    timeoutMapRef.current.clear()
-    intervalMapRef.current.clear()
-  }, [])
-
   // 组件卸载时自动清理所有定时器
   useEffect(() => {
     return () => clearAllTimers()
-  }, [clearAllTimers])
+  }, [])
 
   /**
    * 设置一个 setTimeout 定时器
@@ -93,15 +65,12 @@ export const useTimer = () => {
    * cleanup();
    * ```
    */
-  const setTimeoutTimer = useCallback(
-    (key: string, ...args: Parameters<typeof setTimeout>) => {
-      clearTimeout(timeoutMapRef.current.get(key))
-      const timer = setTimeout(...args)
-      timeoutMapRef.current.set(key, timer)
-      return () => clearTimeoutTimer(key)
-    },
-    [clearTimeoutTimer]
-  )
+  const setTimeoutTimer = (key: string, ...args: Parameters<typeof setTimeout>) => {
+    clearTimeout(timeoutMapRef.current.get(key))
+    const timer = setTimeout(...args)
+    timeoutMapRef.current.set(key, timer)
+    return () => clearTimeoutTimer(key)
+  }
 
   /**
    * 设置一个 setInterval 定时器
@@ -120,31 +89,56 @@ export const useTimer = () => {
    * cleanup();
    * ```
    */
-  const setIntervalTimer = useCallback(
-    (key: string, ...args: Parameters<typeof setInterval>) => {
-      clearInterval(intervalMapRef.current.get(key))
-      const timer = setInterval(...args)
-      intervalMapRef.current.set(key, timer)
-      return () => clearIntervalTimer(key)
-    },
-    [clearIntervalTimer]
-  )
+  const setIntervalTimer = (key: string, ...args: Parameters<typeof setInterval>) => {
+    clearInterval(intervalMapRef.current.get(key))
+    const timer = setInterval(...args)
+    intervalMapRef.current.set(key, timer)
+    return () => clearIntervalTimer(key)
+  }
+
+  /**
+   * 清除指定 key 的 setTimeout 定时器
+   * @param key - 定时器标识符
+   */
+  const clearTimeoutTimer = (key: string) => {
+    clearTimeout(timeoutMapRef.current.get(key))
+    timeoutMapRef.current.delete(key)
+  }
+
+  /**
+   * 清除指定 key 的 setInterval 定时器
+   * @param key - 定时器标识符
+   */
+  const clearIntervalTimer = (key: string) => {
+    clearInterval(intervalMapRef.current.get(key))
+    intervalMapRef.current.delete(key)
+  }
 
   /**
    * 清除所有 setTimeout 定时器
    */
-  const clearAllTimeoutTimers = useCallback(() => {
+  const clearAllTimeoutTimers = () => {
     timeoutMapRef.current.forEach((timer) => clearTimeout(timer))
     timeoutMapRef.current.clear()
-  }, [])
+  }
 
   /**
    * 清除所有 setInterval 定时器
    */
-  const clearAllIntervalTimers = useCallback(() => {
+  const clearAllIntervalTimers = () => {
     intervalMapRef.current.forEach((timer) => clearInterval(timer))
     intervalMapRef.current.clear()
-  }, [])
+  }
+
+  /**
+   * 清除所有定时器，包括 setTimeout 和 setInterval
+   */
+  const clearAllTimers = () => {
+    timeoutMapRef.current.forEach((timer) => clearTimeout(timer))
+    intervalMapRef.current.forEach((timer) => clearInterval(timer))
+    timeoutMapRef.current.clear()
+    intervalMapRef.current.clear()
+  }
 
   return {
     setTimeoutTimer,

src/renderer/src/i18n/locales/en-us.json

@@ -280,7 +280,6 @@
         "denied": "Tool request was denied.",
         "timeout": "Tool request timed out before receiving approval."
       },
-      "toolPendingFallback": "Tool",
       "waiting": "Waiting for tool permission decision..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Image Generation (OpenAI)",
+    "image-generation": "Image Generation",
     "jina-rerank": "Jina Rerank",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} citations",
     "citations": "References",
+    "continue_generation": {
+      "prompt": "[CONTINUE EXACTLY FROM CUTOFF POINT]\n\nYour previous response was cut off mid-generation. Continue IMMEDIATELY from where you stopped - do NOT repeat, summarize, or restart. Your next word should be the exact continuation.\n\nYour response ended with: \"{{truncatedContent}}\"\n\nContinue now (first word must follow directly from the above):"
+    },
     "copied": "Copied!",
     "copy": {
       "failed": "Copy failed",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "Content was blocked by safety filter",
+        "continue": "Continue generating",
+        "error": "An error occurred during generation",
+        "length": "Maximum output length limit reached",
+        "other": "Generation terminated",
+        "unknown": "Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Too many requests. Please wait {{seconds}} seconds before trying again."
       }

src/renderer/src/i18n/locales/zh-cn.json

@@ -280,7 +280,6 @@
         "denied": "工具请求已被拒绝。",
         "timeout": "工具请求在收到批准前超时。"
       },
-      "toolPendingFallback": "工具",
       "waiting": "等待工具权限决定..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "图像生成 (OpenAI)",
+    "image-generation": "图片生成",
     "jina-rerank": "Jina 重排序",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} 个引用内容",
     "citations": "引用内容",
+    "continue_generation": {
+      "prompt": "[从截断处精确继续]\n\n你之前的回复在生成过程中被截断了。请立即从停止的地方继续——不要重复、总结或重新开始。你的下一个字必须是精确的接续。\n\n你的回复结尾是：\"{{truncatedContent}}\"\n\n现在继续（第一个字必须直接接上面的内容）："
+    },
     "copied": "已复制",
     "copy": {
       "failed": "复制失败",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "内容被安全过滤器拦截",
+        "continue": "继续生成",
+        "error": "生成过程中发生错误",
+        "length": "已达到最大输出长度限制",
+        "other": "生成已终止",
+        "unknown": "生成因未知原因终止"
+      },
       "rate": {
         "limit": "发送过于频繁，请等待 {{seconds}} 秒后再尝试"
       }

src/renderer/src/i18n/locales/zh-tw.json

@@ -280,7 +280,6 @@
         "denied": "工具請求已被拒絕。",
         "timeout": "工具請求在收到核准前逾時。"
       },
-      "toolPendingFallback": "工具",
       "waiting": "等待工具權限決定..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "圖像生成 (OpenAI)",
+    "image-generation": "圖片生成",
     "jina-rerank": "Jina Rerank",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} 個引用內容",
     "citations": "引用內容",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "已複製！",
     "copy": {
       "failed": "複製失敗",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "發送過於頻繁，請在 {{seconds}} 秒後再嘗試"
       }

src/renderer/src/i18n/translate/de-de.json

@@ -280,7 +280,6 @@
         "denied": "Tool-Anfrage wurde abgelehnt.",
         "timeout": "Tool-Anfrage ist abgelaufen, bevor eine Genehmigung eingegangen ist."
       },
-      "toolPendingFallback": "Werkzeug",
       "waiting": "Warten auf Entscheidung über Tool-Berechtigung..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Bilderzeugung (OpenAI)",
+    "image-generation": "Bildgenerierung",
     "jina-rerank": "Jina Reranking",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} Zitate",
     "citations": "Zitate",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Kopiert",
     "copy": {
       "failed": "Kopieren fehlgeschlagen",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Zu viele Anfragen. Bitte warten Sie {{seconds}} Sekunden, bevor Sie es erneut versuchen"
       }

src/renderer/src/i18n/translate/el-gr.json

@@ -280,7 +280,6 @@
         "denied": "Το αίτημα για εργαλείο απορρίφθηκε.",
         "timeout": "Το αίτημα για το εργαλείο έληξε πριν λάβει έγκριση."
       },
-      "toolPendingFallback": "Εργαλείο",
       "waiting": "Αναμονή για απόφαση άδειας εργαλείου..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Δημιουργία Εικόνων (OpenAI)",
+    "image-generation": "Δημιουργία Εικόνας",
     "jina-rerank": "Επαναταξινόμηση Jina",
     "openai": "OpenAI",
     "openai-response": "Απάντηση OpenAI"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} αναφορές",
     "citations": "Περιεχόμενα αναφοράς",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Αντιγράφηκε",
     "copy": {
       "failed": "Η αντιγραφή απέτυχε",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Υπερβολική συχνότητα στείλατε παρακαλώ περιμένετε {{seconds}} δευτερόλεπτα και προσπαθήστε ξανά"
       }

src/renderer/src/i18n/translate/es-es.json

@@ -280,7 +280,6 @@
         "denied": "La solicitud de herramienta fue denegada.",
         "timeout": "La solicitud de herramienta expiró antes de recibir la aprobación."
       },
-      "toolPendingFallback": "Herramienta",
       "waiting": "Esperando la decisión de permiso de la herramienta..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Generación de Imágenes (OpenAI)",
+    "image-generation": "Generación de imágenes",
     "jina-rerank": "Reordenamiento Jina",
     "openai": "OpenAI",
     "openai-response": "Respuesta de OpenAI"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} contenido citado",
     "citations": "Citas",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Copiado",
     "copy": {
       "failed": "Copia fallida",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Envío demasiado frecuente, espere {{seconds}} segundos antes de intentarlo de nuevo"
       }

src/renderer/src/i18n/translate/fr-fr.json

@@ -280,7 +280,6 @@
         "denied": "La demande d'outil a été refusée.",
         "timeout": "La demande d'outil a expiré avant d'obtenir l'approbation."
       },
-      "toolPendingFallback": "Outil",
       "waiting": "En attente de la décision d'autorisation de l'outil..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Génération d'images (OpenAI)",
+    "image-generation": "Génération d'images",
     "jina-rerank": "Reclassement Jina",
     "openai": "OpenAI",
     "openai-response": "Réponse OpenAI"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} éléments cités",
     "citations": "Citations",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Copié",
     "copy": {
       "failed": "La copie a échoué",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Vous envoyez trop souvent, veuillez attendre {{seconds}} secondes avant de réessayer"
       }

src/renderer/src/i18n/translate/ja-jp.json

@@ -280,7 +280,6 @@
         "denied": "ツールリクエストは拒否されました。",
         "timeout": "ツールリクエストは承認を受ける前にタイムアウトしました。"
       },
-      "toolPendingFallback": "ツール",
       "waiting": "ツールの許可決定を待っています..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "画像生成 (OpenAI)",
+    "image-generation": "画像生成",
     "jina-rerank": "Jina Rerank",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}}個の引用内容",
     "citations": "引用内容",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "コピーしました！",
     "copy": {
       "failed": "コピーに失敗しました",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "送信が頻繁すぎます。{{seconds}} 秒待ってから再試行してください。"
       }

src/renderer/src/i18n/translate/pt-pt.json

@@ -16,7 +16,7 @@
       "error": {
         "failed": "Falha ao excluir o agente"
       },
-      "title": "Excluir Agente"
+      "title": "删除代理"
     },
     "edit": {
       "title": "Agent Editor"
@@ -111,7 +111,7 @@
           "label": "Modo de permissão",
           "options": {
             "acceptEdits": "Aceitar edições automaticamente",
-            "bypassPermissions": "Ignorar verificações de permissão",
+            "bypassPermissions": "忽略检查 de permissão",
             "default": "Padrão (perguntar antes de continuar)",
             "plan": "Modo de planejamento (plano sujeito a aprovação)"
           },
@@ -150,7 +150,7 @@
         },
         "success": {
           "install": "Plugin instalado com sucesso",
-          "uninstall": "Plugin desinstalado com sucesso"
+          "uninstall": "插件 desinstalado com sucesso"
         },
         "tab": "plug-in",
         "type": {
@@ -280,7 +280,6 @@
         "denied": "Solicitação de ferramenta foi negada.",
         "timeout": "A solicitação da ferramenta expirou antes de receber aprovação."
       },
-      "toolPendingFallback": "Ferramenta",
       "waiting": "Aguardando decisão de permissão da ferramenta..."
     },
     "type": {
@@ -1135,7 +1134,7 @@
     "duplicate": "Duplicar",
     "edit": "Editar",
     "enabled": "Ativado",
-    "error": "Erro",
+    "error": "错误",
     "errors": {
       "create_message": "Falha ao criar mensagem",
       "validation": "Falha na verificação"
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Geração de Imagens (OpenAI)",
+    "image-generation": "Geração de Imagem",
     "jina-rerank": "Jina Reordenar",
     "openai": "OpenAI",
     "openai-response": "Resposta OpenAI"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} conteúdo(s) citado(s)",
     "citations": "Citações",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Copiado",
     "copy": {
       "failed": "Cópia falhou",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Envio muito frequente, aguarde {{seconds}} segundos antes de tentar novamente"
       }

src/renderer/src/i18n/translate/ru-ru.json

@@ -280,7 +280,6 @@
         "denied": "Запрос на инструмент был отклонён.",
         "timeout": "Запрос на инструмент превысил время ожидания до получения подтверждения."
       },
-      "toolPendingFallback": "Инструмент",
       "waiting": "Ожидание решения о разрешении на использование инструмента..."
     },
     "type": {
@@ -1209,7 +1208,7 @@
   "endpoint_type": {
     "anthropic": "Anthropic",
     "gemini": "Gemini",
-    "image-generation": "Генерация изображений (OpenAI)",
+    "image-generation": "Изображение",
     "jina-rerank": "Jina Rerank",
     "openai": "OpenAI",
     "openai-response": "OpenAI-Response"
@@ -1731,6 +1730,9 @@
     },
     "citation": "{{count}} цитат",
     "citations": "Содержание цитат",
+    "continue_generation": {
+      "prompt": "[to be translated]:Please continue your previous response exactly from where you left off. Do not repeat any content that was already generated. Continue directly from:\n\n{{truncatedContent}}"
+    },
     "copied": "Скопировано!",
     "copy": {
       "failed": "Не удалось скопировать",
@@ -1942,6 +1944,14 @@
       }
     },
     "warning": {
+      "finish_reason": {
+        "content-filter": "[to be translated]:Content was blocked by safety filter",
+        "continue": "[to be translated]:Continue generating",
+        "error": "[to be translated]:An error occurred during generation",
+        "length": "[to be translated]:Maximum output length limit reached",
+        "other": "[to be translated]:Generation terminated",
+        "unknown": "[to be translated]:Generation terminated for unknown reason"
+      },
       "rate": {
         "limit": "Отправка слишком частая, пожалуйста, подождите {{seconds}} секунд, прежде чем попробовать снова."
       }

src/renderer/src/pages/home/Inputbar/AgentSessionInputbar.tsx

@@ -9,7 +9,6 @@ import { getModel } from '@renderer/hooks/useModel'
 import { useSettings } from '@renderer/hooks/useSettings'
 import { useTextareaResize } from '@renderer/hooks/useTextareaResize'
 import { useTimer } from '@renderer/hooks/useTimer'
-import { CacheService } from '@renderer/services/CacheService'
 import { pauseTrace } from '@renderer/services/SpanManagerService'
 import { estimateUserPromptUsage } from '@renderer/services/TokenService'
 import { useAppDispatch, useAppSelector } from '@renderer/store'
@@ -42,10 +41,19 @@ import { getInputbarConfig } from './registry'
 import { TopicType } from './types'
 
 const logger = loggerService.withContext('AgentSessionInputbar')
+const agentSessionDraftCache = new Map<string, string>()
 
-const DRAFT_CACHE_TTL = 24 * 60 * 60 * 1000 // 24 hours
+const readDraftFromCache = (key: string): string => {
+  return agentSessionDraftCache.get(key) ?? ''
+}
 
-const getAgentDraftCacheKey = (agentId: string) => `agent-session-draft-${agentId}`
+const writeDraftToCache = (key: string, value: string) => {
+  if (!value) {
+    agentSessionDraftCache.delete(key)
+  } else {
+    agentSessionDraftCache.set(key, value)
+  }
+}
 
 type Props = {
   agentId: string
@@ -162,15 +170,16 @@ const AgentSessionInputbarInner: FC<InnerProps> = ({ assistant, agentId, session
   const scope = TopicType.Session
   const config = getInputbarConfig(scope)
 
-  // Use shared hooks for text and textarea management with draft persistence
-  const draftCacheKey = getAgentDraftCacheKey(agentId)
+  // Use shared hooks for text and textarea management
+  const initialDraft = useMemo(() => readDraftFromCache(agentId), [agentId])
+  const persistDraft = useCallback((next: string) => writeDraftToCache(agentId, next), [agentId])
   const {
     text,
     setText,
     isEmpty: inputEmpty
   } = useInputText({
-    initialValue: CacheService.get<string>(draftCacheKey) ?? '',
-    onChange: (value) => CacheService.set(draftCacheKey, value, DRAFT_CACHE_TTL)
+    initialValue: initialDraft,
+    onChange: persistDraft
   })
   const {
     textareaRef,
@@ -422,7 +431,6 @@ const AgentSessionInputbarInner: FC<InnerProps> = ({ assistant, agentId, session
         })
       )
 
-      // Clear text after successful send (draft is cleared automatically via onChange)
       setText('')
       setTimeoutTimer('agentSession_sendMessage', () => setText(''), 500)
     } catch (error) {

src/renderer/src/pages/home/Inputbar/Inputbar.tsx

@@ -14,6 +14,7 @@ import { useInputText } from '@renderer/hooks/useInputText'
 import { useMessageOperations, useTopicLoading } from '@renderer/hooks/useMessageOperations'
 import { useSettings } from '@renderer/hooks/useSettings'
 import { useShortcut } from '@renderer/hooks/useShortcuts'
+import { useSidebarIconShow } from '@renderer/hooks/useSidebarIcon'
 import { useTextareaResize } from '@renderer/hooks/useTextareaResize'
 import { useTimer } from '@renderer/hooks/useTimer'
 import {
@@ -23,7 +24,6 @@ import {
   useInputbarToolsState
 } from '@renderer/pages/home/Inputbar/context/InputbarToolsProvider'
 import { getDefaultTopic } from '@renderer/services/AssistantService'
-import { CacheService } from '@renderer/services/CacheService'
 import { EVENT_NAMES, EventEmitter } from '@renderer/services/EventService'
 import FileManager from '@renderer/services/FileManager'
 import { checkRateLimit, getUserMessage } from '@renderer/services/MessagesService'
@@ -39,7 +39,7 @@ import { getSendMessageShortcutLabel } from '@renderer/utils/input'
 import { documentExts, imageExts, textExts } from '@shared/config/constant'
 import { debounce } from 'lodash'
 import type { FC } from 'react'
-import React, { useCallback, useEffect, useEffectEvent, useMemo, useRef, useState } from 'react'
+import React, { useCallback, useEffect, useMemo, useRef, useState } from 'react'
 import { useTranslation } from 'react-i18next'
 
 import { InputbarCore } from './components/InputbarCore'
@@ -51,17 +51,6 @@ import TokenCount from './TokenCount'
 
 const logger = loggerService.withContext('Inputbar')
 
-const INPUTBAR_DRAFT_CACHE_KEY = 'inputbar-draft'
-const DRAFT_CACHE_TTL = 24 * 60 * 60 * 1000 // 24 hours
-
-const getMentionedModelsCacheKey = (assistantId: string) => `inputbar-mentioned-models-${assistantId}`
-
-const getValidatedCachedModels = (assistantId: string): Model[] => {
-  const cached = CacheService.get<Model[]>(getMentionedModelsCacheKey(assistantId))
-  if (!Array.isArray(cached)) return []
-  return cached.filter((model) => model?.id && model?.name)
-}
-
 interface Props {
   assistant: Assistant
   setActiveTopic: (topic: Topic) => void
@@ -91,18 +80,16 @@ const Inputbar: FC<Props> = ({ assistant: initialAssistant, setActiveTopic, topi
     toggleExpanded: () => {}
   })
 
-  const [initialMentionedModels] = useState(() => getValidatedCachedModels(initialAssistant.id))
-
   const initialState = useMemo(
     () => ({
       files: [] as FileType[],
-      mentionedModels: initialMentionedModels,
+      mentionedModels: [] as Model[],
       selectedKnowledgeBases: initialAssistant.knowledge_bases ?? [],
       isExpanded: false,
       couldAddImageFile: false,
       extensions: [] as string[]
     }),
-    [initialMentionedModels, initialAssistant.knowledge_bases]
+    [initialAssistant.knowledge_bases]
   )
 
   return (
@@ -134,10 +121,7 @@ const InputbarInner: FC<InputbarInnerProps> = ({ assistant: initialAssistant, se
   const { setFiles, setMentionedModels, setSelectedKnowledgeBases } = useInputbarToolsDispatch()
   const { setCouldAddImageFile } = useInputbarToolsInternalDispatch()
 
-  const { text, setText } = useInputText({
-    initialValue: CacheService.get<string>(INPUTBAR_DRAFT_CACHE_KEY) ?? '',
-    onChange: (value) => CacheService.set(INPUTBAR_DRAFT_CACHE_KEY, value, DRAFT_CACHE_TTL)
-  })
+  const { text, setText } = useInputText()
   const {
     textareaRef,
     resize: resizeTextArea,
@@ -149,6 +133,7 @@ const InputbarInner: FC<InputbarInnerProps> = ({ assistant: initialAssistant, se
     minHeight: 30
   })
 
+  const showKnowledgeIcon = useSidebarIconShow('knowledge')
   const { assistant, addTopic, model, setModel, updateAssistant } = useAssistant(initialAssistant.id)
   const { sendMessageShortcut, showInputEstimatedTokens, enableQuickPanelTriggers } = useSettings()
   const [estimateTokenCount, setEstimateTokenCount] = useState(0)
@@ -205,15 +190,6 @@ const InputbarInner: FC<InputbarInnerProps> = ({ assistant: initialAssistant, se
     setCouldAddImageFile(canAddImageFile)
   }, [canAddImageFile, setCouldAddImageFile])
 
-  const onUnmount = useEffectEvent((id: string) => {
-    CacheService.set(getMentionedModelsCacheKey(id), mentionedModels, DRAFT_CACHE_TTL)
-  })
-
-  useEffect(() => {
-    return () => onUnmount(assistant.id)
-    // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [assistant.id])
-
   const placeholderText = enableQuickPanelTriggers
     ? t('chat.input.placeholder', { key: getSendMessageShortcutLabel(sendMessageShortcut) })
     : t('chat.input.placeholder_without_triggers', {
@@ -405,10 +381,9 @@ const InputbarInner: FC<InputbarInnerProps> = ({ assistant: initialAssistant, se
     focusTextarea
   ])
 
-  // TODO: Just use assistant.knowledge_bases as selectedKnowledgeBases. context state is overdesigned.
   useEffect(() => {
-    setSelectedKnowledgeBases(assistant.knowledge_bases ?? [])
-  }, [assistant.knowledge_bases, setSelectedKnowledgeBases])
+    setSelectedKnowledgeBases(showKnowledgeIcon ? (assistant.knowledge_bases ?? []) : [])
+  }, [assistant.knowledge_bases, setSelectedKnowledgeBases, showKnowledgeIcon])
 
   useEffect(() => {
     // Disable web search if model doesn't support it

src/renderer/src/pages/home/Inputbar/components/InputbarCore.tsx

@@ -156,8 +156,11 @@ export const InputbarCore: FC<InputbarCoreProps> = ({
 
   const setText = useCallback<React.Dispatch<React.SetStateAction<string>>>(
     (value) => {
-      const newText = typeof value === 'function' ? value(textRef.current) : value
-      onTextChange(newText)
+      if (typeof value === 'function') {
+        onTextChange(value(textRef.current))
+      } else {
+        onTextChange(value)
+      }
     },
     [onTextChange]
   )

src/renderer/src/pages/home/Inputbar/tools/knowledgeBaseTool.tsx

@@ -1,4 +1,5 @@
 import { useAssistant } from '@renderer/hooks/useAssistant'
+import { useSidebarIconShow } from '@renderer/hooks/useSidebarIcon'
 import { defineTool, registerTool, TopicType } from '@renderer/pages/home/Inputbar/types'
 import type { KnowledgeBase } from '@renderer/types'
 import { isPromptToolUse, isSupportedToolUse } from '@renderer/utils/mcp-tools'
@@ -29,6 +30,7 @@ const knowledgeBaseTool = defineTool({
   render: function KnowledgeBaseToolRender(context) {
     const { assistant, state, actions, quickPanel } = context
 
+    const knowledgeSidebarEnabled = useSidebarIconShow('knowledge')
     const { updateAssistant } = useAssistant(assistant.id)
 
     const handleSelect = useCallback(
@@ -39,6 +41,10 @@ const knowledgeBaseTool = defineTool({
       [updateAssistant, actions]
     )
 
+    if (!knowledgeSidebarEnabled) {
+      return null
+    }
+
     return (
       <KnowledgeBaseButton
         quickPanel={quickPanel}

src/renderer/src/pages/home/Messages/Blocks/ThinkingBlock.tsx

@@ -102,12 +102,10 @@ const ThinkingBlock: React.FC<Props> = ({ block }) => {
   )
 }
 
-const normalizeThinkingTime = (value?: number) => (typeof value === 'number' && Number.isFinite(value) ? value : 0)
-
 const ThinkingTimeSeconds = memo(
   ({ blockThinkingTime, isThinking }: { blockThinkingTime: number; isThinking: boolean }) => {
     const { t } = useTranslation()
-    const [displayTime, setDisplayTime] = useState(normalizeThinkingTime(blockThinkingTime))
+    const [displayTime, setDisplayTime] = useState(blockThinkingTime)
 
     const timer = useRef<NodeJS.Timeout | null>(null)
 
@@ -123,7 +121,7 @@ const ThinkingTimeSeconds = memo(
           clearInterval(timer.current)
           timer.current = null
         }
-        setDisplayTime(normalizeThinkingTime(blockThinkingTime))
+        setDisplayTime(blockThinkingTime)
       }
 
       return () => {
@@ -134,10 +132,10 @@ const ThinkingTimeSeconds = memo(
       }
     }, [isThinking, blockThinkingTime])
 
-    const thinkingTimeSeconds = useMemo(() => {
-      const safeTime = normalizeThinkingTime(displayTime)
-      return ((safeTime < 1000 ? 100 : safeTime) / 1000).toFixed(1)
-    }, [displayTime])
+    const thinkingTimeSeconds = useMemo(
+      () => ((displayTime < 1000 ? 100 : displayTime) / 1000).toFixed(1),
+      [displayTime]
+    )
 
     return isThinking
       ? t('chat.thinking', {

src/renderer/src/pages/home/Messages/Blocks/__tests__/ThinkingBlock.test.tsx

@@ -255,20 +255,6 @@ describe('ThinkingBlock', () => {
         unmount()
       })
     })
-
-    it('should clamp invalid thinking times to a safe default', () => {
-      const testCases = [undefined, Number.NaN, Number.POSITIVE_INFINITY]
-
-      testCases.forEach((thinking_millsec) => {
-        const block = createThinkingBlock({
-          thinking_millsec: thinking_millsec as any,
-          status: MessageBlockStatus.SUCCESS
-        })
-        const { unmount } = renderThinkingBlock(block)
-        expect(getThinkingTimeText()).toHaveTextContent('0.1s')
-        unmount()
-      })
-    })
   })
 
   describe('collapse behavior', () => {

src/renderer/src/pages/home/Messages/ChatNavigation.tsx

@@ -10,7 +10,6 @@ import { useSettings } from '@renderer/hooks/useSettings'
 import { useTimer } from '@renderer/hooks/useTimer'
 import type { RootState } from '@renderer/store'
 // import { selectCurrentTopicId } from '@renderer/store/newMessage'
-import { scrollIntoView } from '@renderer/utils/dom'
 import { Button, Drawer, Tooltip } from 'antd'
 import type { FC } from 'react'
 import { useCallback, useEffect, useRef, useState } from 'react'
@@ -119,8 +118,7 @@ const ChatNavigation: FC<ChatNavigationProps> = ({ containerId }) => {
   }
 
   const scrollToMessage = (element: HTMLElement) => {
-    // Use container: 'nearest' to keep scroll within the chat pane (Chromium-only, see #11565, #11567)
-    scrollIntoView(element, { behavior: 'smooth', block: 'start', container: 'nearest' })
+    element.scrollIntoView({ behavior: 'smooth', block: 'start' })
   }
 
   const scrollToTop = () => {

src/renderer/src/pages/home/Messages/FinishReasonWarning.tsx

@@ -0,0 +1,64 @@
+import type { FinishReason } from 'ai'
+import { Alert as AntdAlert, Button } from 'antd'
+import { Play } from 'lucide-react'
+import React from 'react'
+import { useTranslation } from 'react-i18next'
+import styled from 'styled-components'
+
+interface Props {
+  finishReason: FinishReason
+  onContinue?: () => void
+  onDismiss?: () => void
+}
+
+/**
+ * Displays a warning banner when message generation was truncated or filtered
+ * Only shows for non-normal finish reasons (not 'stop' or 'tool-calls')
+ */
+const FinishReasonWarning: React.FC<Props> = ({ finishReason, onContinue, onDismiss }) => {
+  const { t } = useTranslation()
+
+  // Don't show warning for normal finish reasons
+  if (finishReason === 'stop' || finishReason === 'tool-calls') {
+    return null
+  }
+
+  const getWarningMessage = () => {
+    const i18nKey = `message.warning.finish_reason.${finishReason}`
+    return t(i18nKey)
+  }
+
+  // Only show continue button for 'length' reason (max tokens reached)
+  const showContinueButton = finishReason === 'length' && onContinue
+
+  return (
+    <Alert
+      message={getWarningMessage()}
+      type="warning"
+      showIcon
+      closable={!!onDismiss}
+      onClose={onDismiss}
+      action={
+        showContinueButton && (
+          <Button
+            size="small"
+            type="text"
+            icon={<Play size={14} />}
+            onClick={onContinue}
+            style={{ display: 'flex', alignItems: 'center', gap: 4 }}>
+            {t('message.warning.finish_reason.continue')}
+          </Button>
+        )
+      }
+    />
+  )
+}
+
+const Alert = styled(AntdAlert)`
+  margin: 0.5rem 0 !important;
+  padding: 8px 12px;
+  font-size: 12px;
+  align-items: center;
+`
+
+export default React.memo(FinishReasonWarning)

src/renderer/src/pages/home/Messages/Message.tsx

@@ -13,9 +13,8 @@ import { getMessageModelId } from '@renderer/services/MessagesService'
 import { getModelUniqId } from '@renderer/services/ModelService'
 import { estimateMessageUsage } from '@renderer/services/TokenService'
 import type { Assistant, Topic } from '@renderer/types'
-import type { Message, MessageBlock } from '@renderer/types/newMessage'
+import type { Message as MessageType, MessageBlock } from '@renderer/types/newMessage'
 import { classNames, cn } from '@renderer/utils'
-import { scrollIntoView } from '@renderer/utils/dom'
 import { isMessageProcessing } from '@renderer/utils/messageUtils/is'
 import { Divider } from 'antd'
 import type { Dispatch, FC, SetStateAction } from 'react'
@@ -31,7 +30,7 @@ import MessageMenubar from './MessageMenubar'
 import MessageOutline from './MessageOutline'
 
 interface Props {
-  message: Message
+  message: MessageType
   topic: Topic
   assistant?: Assistant
   index?: number
@@ -40,7 +39,7 @@ interface Props {
   style?: React.CSSProperties
   isGrouped?: boolean
   isStreaming?: boolean
-  onSetMessages?: Dispatch<SetStateAction<Message[]>>
+  onSetMessages?: Dispatch<SetStateAction<MessageType[]>>
   onUpdateUseful?: (msgId: string) => void
   isGroupContextMessage?: boolean
 }
@@ -80,10 +79,9 @@ const MessageItem: FC<Props> = ({
 
   useEffect(() => {
     if (isEditing && messageContainerRef.current) {
-      scrollIntoView(messageContainerRef.current, {
+      messageContainerRef.current.scrollIntoView({
         behavior: 'smooth',
-        block: 'center',
-        container: 'nearest'
+        block: 'center'
       })
     }
   }, [isEditing])
@@ -118,6 +116,26 @@ const MessageItem: FC<Props> = ({
     stopEditing()
   }, [stopEditing])
 
+  // Handle continue generation when max tokens reached
+  const handleContinueGeneration = useCallback(
+    async (msg: MessageType) => {
+      if (!assistant) return
+      // Clear the finishReason first, then trigger continue generation
+      await editMessage(msg.id, { finishReason: undefined })
+      // Emit event to trigger continue generation
+      EventEmitter.emit(EVENT_NAMES.CONTINUE_GENERATION, { message: msg, assistant, topic })
+    },
+    [assistant, editMessage, topic]
+  )
+
+  // Handle dismiss warning (just clear finishReason)
+  const handleDismissWarning = useCallback(
+    async (msg: MessageType) => {
+      await editMessage(msg.id, { finishReason: undefined })
+    },
+    [editMessage]
+  )
+
   const isLastMessage = index === 0 || !!isGrouped
   const isAssistantMessage = message.role === 'assistant'
   const isProcessing = isMessageProcessing(message)
@@ -126,7 +144,7 @@ const MessageItem: FC<Props> = ({
   const messageHighlightHandler = useCallback(
     (highlight: boolean = true) => {
       if (messageContainerRef.current) {
-        scrollIntoView(messageContainerRef.current, { behavior: 'smooth', block: 'center', container: 'nearest' })
+        messageContainerRef.current.scrollIntoView({ behavior: 'smooth' })
         if (highlight) {
           setTimeoutTimer(
             'messageHighlightHandler',
@@ -225,7 +243,11 @@ const MessageItem: FC<Props> = ({
                 overflowY: 'visible'
               }}>
               <MessageErrorBoundary>
-                <MessageContent message={message} />
+                <MessageContent
+                  message={message}
+                  onContinueGeneration={handleContinueGeneration}
+                  onDismissWarning={handleDismissWarning}
+                />
               </MessageErrorBoundary>
             </MessageContentContainer>
             {showMenubar && (

src/renderer/src/pages/home/Messages/MessageAnchorLine.tsx

@@ -12,7 +12,6 @@ import { newMessagesActions } from '@renderer/store/newMessage'
 // import { updateMessageThunk } from '@renderer/store/thunk/messageThunk'
 import type { Message } from '@renderer/types/newMessage'
 import { isEmoji, removeLeadingEmoji } from '@renderer/utils'
-import { scrollIntoView } from '@renderer/utils/dom'
 import { getMainTextContent } from '@renderer/utils/messageUtils/find'
 import { Avatar } from 'antd'
 import { CircleChevronDown } from 'lucide-react'
@@ -120,7 +119,7 @@ const MessageAnchorLine: FC<MessageLineProps> = ({ messages }) => {
           () => {
             const messageElement = document.getElementById(`message-${message.id}`)
             if (messageElement) {
-              scrollIntoView(messageElement, { behavior: 'auto', block: 'start', container: 'nearest' })
+              messageElement.scrollIntoView({ behavior: 'auto', block: 'start' })
             }
           },
           100
@@ -142,7 +141,7 @@ const MessageAnchorLine: FC<MessageLineProps> = ({ messages }) => {
         return
       }
 
-      scrollIntoView(messageElement, { behavior: 'smooth', block: 'start', container: 'nearest' })
+      messageElement.scrollIntoView({ behavior: 'smooth', block: 'start' })
     },
     [setSelectedMessage]
   )

src/renderer/src/pages/home/Messages/MessageContent.tsx

@@ -6,11 +6,29 @@ import React from 'react'
 import styled from 'styled-components'
 
 import MessageBlockRenderer from './Blocks'
+import FinishReasonWarning from './FinishReasonWarning'
+
 interface Props {
   message: Message
+  onContinueGeneration?: (message: Message) => void
+  onDismissWarning?: (message: Message) => void
 }
 
-const MessageContent: React.FC<Props> = ({ message }) => {
+const MessageContent: React.FC<Props> = ({ message, onContinueGeneration, onDismissWarning }) => {
+  // Check if we should show finish reason warning
+  const showFinishReasonWarning =
+    message.role === 'assistant' &&
+    message.finishReason &&
+    !['stop', 'tool-calls', 'error'].includes(message.finishReason)
+
+  const handleContinue = () => {
+    onContinueGeneration?.(message)
+  }
+
+  const handleDismiss = () => {
+    onDismissWarning?.(message)
+  }
+
   return (
     <>
       {!isEmpty(message.mentions) && (
@@ -21,6 +39,13 @@ const MessageContent: React.FC<Props> = ({ message }) => {
         </Flex>
       )}
       <MessageBlockRenderer blocks={message.blocks} message={message} />
+      {showFinishReasonWarning && (
+        <FinishReasonWarning
+          finishReason={message.finishReason!}
+          onContinue={onContinueGeneration ? handleContinue : undefined}
+          onDismiss={onDismissWarning ? handleDismiss : undefined}
+        />
+      )}
     </>
   )
 }

src/renderer/src/pages/home/Messages/MessageGroup.tsx

@@ -10,7 +10,6 @@ import type { MultiModelMessageStyle } from '@renderer/store/settings'
 import type { Topic } from '@renderer/types'
 import type { Message } from '@renderer/types/newMessage'
 import { classNames } from '@renderer/utils'
-import { scrollIntoView } from '@renderer/utils/dom'
 import { Popover } from 'antd'
 import type { ComponentProps } from 'react'
 import { memo, useCallback, useEffect, useMemo, useState } from 'react'
@@ -74,7 +73,7 @@ const MessageGroup = ({ messages, topic, registerMessageElement }: Props) => {
         () => {
           const messageElement = document.getElementById(`message-${message.id}`)
           if (messageElement) {
-            scrollIntoView(messageElement, { behavior: 'smooth', block: 'start', container: 'nearest' })
+            messageElement.scrollIntoView({ behavior: 'smooth', block: 'start' })
           }
         },
         200
@@ -133,7 +132,7 @@ const MessageGroup = ({ messages, topic, registerMessageElement }: Props) => {
             setSelectedMessage(message)
           } else {
             // 直接滚动
-            scrollIntoView(element, { behavior: 'smooth', block: 'start', container: 'nearest' })
+            element.scrollIntoView({ behavior: 'smooth', block: 'start' })
           }
         }
       }

src/renderer/src/pages/home/Messages/MessageOutline.tsx

@@ -3,7 +3,6 @@ import type { RootState } from '@renderer/store'
 import { messageBlocksSelectors } from '@renderer/store/messageBlock'
 import type { Message } from '@renderer/types/newMessage'
 import { MessageBlockType } from '@renderer/types/newMessage'
-import { scrollIntoView } from '@renderer/utils/dom'
 import type { FC } from 'react'
 import React, { useMemo, useRef } from 'react'
 import { useSelector } from 'react-redux'
@@ -73,10 +72,10 @@ const MessageOutline: FC<MessageOutlineProps> = ({ message }) => {
     const parent = messageOutlineContainerRef.current?.parentElement
     const messageContentContainer = parent?.querySelector('.message-content-container')
     if (messageContentContainer) {
-      const headingElement = messageContentContainer.querySelector<HTMLElement>(`#${id}`)
+      const headingElement = messageContentContainer.querySelector(`#${id}`)
       if (headingElement) {
         const scrollBlock = ['horizontal', 'grid'].includes(message.multiModelMessageStyle ?? '') ? 'nearest' : 'start'
-        scrollIntoView(headingElement, { behavior: 'smooth', block: scrollBlock, container: 'nearest' })
+        headingElement.scrollIntoView({ behavior: 'smooth', block: scrollBlock })
       }
     }
   }

src/renderer/src/pages/home/Messages/Messages.tsx

@@ -18,7 +18,11 @@ import { estimateHistoryTokens } from '@renderer/services/TokenService'
 import store, { useAppDispatch } from '@renderer/store'
 import { messageBlocksSelectors, updateOneBlock } from '@renderer/store/messageBlock'
 import { newMessagesActions } from '@renderer/store/newMessage'
-import { saveMessageAndBlocksToDB, updateMessageAndBlocksThunk } from '@renderer/store/thunk/messageThunk'
+import {
+  continueGenerationThunk,
+  saveMessageAndBlocksToDB,
+  updateMessageAndBlocksThunk
+} from '@renderer/store/thunk/messageThunk'
 import type { Assistant, Topic } from '@renderer/types'
 import type { MessageBlock } from '@renderer/types/newMessage'
 import { type Message, MessageBlockType } from '@renderer/types/newMessage'
@@ -233,6 +237,18 @@ const Messages: React.FC<MessagesProps> = ({ assistant, topic, setActiveTopic, o
             window.toast.error(t('code_block.edit.save.failed.label'))
           }
         }
+      ),
+      EventEmitter.on(
+        EVENT_NAMES.CONTINUE_GENERATION,
+        async (data: { message: Message; assistant: Assistant; topic: Topic }) => {
+          const { message, assistant: msgAssistant, topic: msgTopic } = data
+          // Only handle if it's for the current topic
+          if (msgTopic.id !== topic.id) {
+            return
+          }
+          await dispatch(continueGenerationThunk(topic.id, message, msgAssistant))
+          scrollToBottom()
+        }
       )
     ]
 

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/BashOutputTool.tsx

@@ -76,7 +76,7 @@ export function BashOutputTool({
   input,
   output
 }: {
-  input?: BashOutputToolInput
+  input: BashOutputToolInput
   output?: BashOutputToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   const parsedOutput = parseBashOutput(output)
@@ -144,7 +144,7 @@ export function BashOutputTool({
           label="Bash Output"
           params={
             <div className="flex items-center gap-2">
-              <Tag className="py-0 font-mono text-xs">{input?.bash_id}</Tag>
+              <Tag className="py-0 font-mono text-xs">{input.bash_id}</Tag>
               {statusConfig && (
                 <Tag
                   color={statusConfig.color}

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/BashTool.tsx

@@ -5,20 +5,24 @@ import { Terminal } from 'lucide-react'
 import { ToolTitle } from './GenericTools'
 import type { BashToolInput as BashToolInputType, BashToolOutput as BashToolOutputType } from './types'
 
+const MAX_TAG_LENGTH = 100
+
 export function BashTool({
   input,
   output
 }: {
-  input?: BashToolInputType
+  input: BashToolInputType
   output?: BashToolOutputType
 }): NonNullable<CollapseProps['items']>[number] {
   // 如果有输出，计算输出行数
   const outputLines = output ? output.split('\n').length : 0
 
-  // 处理命令字符串，添加空值检查
-  const command = input?.command ?? ''
+  // 处理命令字符串的截断
+  const command = input.command
+  const needsTruncate = command.length > MAX_TAG_LENGTH
+  const displayCommand = needsTruncate ? `${command.slice(0, MAX_TAG_LENGTH)}...` : command
 
-  const tagContent = <Tag className="!m-0 max-w-full truncate font-mono">{command}</Tag>
+  const tagContent = <Tag className="whitespace-pre-wrap break-all font-mono">{displayCommand}</Tag>
 
   return {
     key: 'tool',
@@ -27,15 +31,19 @@ export function BashTool({
         <ToolTitle
           icon={<Terminal className="h-4 w-4" />}
           label="Bash"
-          params={input?.description}
+          params={input.description}
           stats={output ? `${outputLines} ${outputLines === 1 ? 'line' : 'lines'}` : undefined}
         />
-        <div className="mt-1 max-w-full">
-          <Popover
-            content={<div className="max-w-xl whitespace-pre-wrap break-all font-mono text-xs">{command}</div>}
-            trigger="hover">
-            {tagContent}
-          </Popover>
+        <div className="mt-1">
+          {needsTruncate ? (
+            <Popover
+              content={<div className="max-w-xl whitespace-pre-wrap break-all font-mono">{command}</div>}
+              trigger="hover">
+              {tagContent}
+            </Popover>
+          ) : (
+            tagContent
+          )}
         </div>
       </>
     ),

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/EditTool.tsx

@@ -32,19 +32,19 @@ export function EditTool({
   input,
   output
 }: {
-  input?: EditToolInput
+  input: EditToolInput
   output?: EditToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   return {
     key: AgentToolsType.Edit,
-    label: <ToolTitle icon={<FileEdit className="h-4 w-4" />} label="Edit" params={input?.file_path} />,
+    label: <ToolTitle icon={<FileEdit className="h-4 w-4" />} label="Edit" params={input.file_path} />,
     children: (
       <>
         {/* Diff View */}
         {/* Old Content */}
-        {renderCodeBlock(input?.old_string ?? '', 'old')}
+        {renderCodeBlock(input.old_string, 'old')}
         {/* New Content */}
-        {renderCodeBlock(input?.new_string ?? '', 'new')}
+        {renderCodeBlock(input.new_string, 'new')}
         {/* Output */}
         {output}
       </>

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/ExitPlanModeTool.tsx

@@ -10,19 +10,18 @@ export function ExitPlanModeTool({
   input,
   output
 }: {
-  input?: ExitPlanModeToolInput
+  input: ExitPlanModeToolInput
   output?: ExitPlanModeToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
-  const plan = input?.plan ?? ''
   return {
     key: AgentToolsType.ExitPlanMode,
     label: (
       <ToolTitle
         icon={<DoorOpen className="h-4 w-4" />}
         label="ExitPlanMode"
-        stats={`${plan.split('\n\n').length} plans`}
+        stats={`${input.plan.split('\n\n').length} plans`}
       />
     ),
-    children: <ReactMarkdown>{plan + '\n\n' + (output ?? '')}</ReactMarkdown>
+    children: <ReactMarkdown>{input.plan + '\n\n' + (output ?? '')}</ReactMarkdown>
   }
 }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/GenericTools.tsx

@@ -18,9 +18,9 @@ export function ToolTitle({
 }) {
   return (
     <div className={`flex items-center gap-1 ${className}`}>
-      {icon && <span className="flex flex-shrink-0">{icon}</span>}
-      {label && <span className="flex-shrink-0 font-medium text-sm">{label}</span>}
-      {params && <span className="min-w-0 truncate text-muted-foreground text-xs">{params}</span>}
+      {icon}
+      {label && <span className="font-medium text-sm">{label}</span>}
+      {params && <span className="flex-shrink-0 text-muted-foreground text-xs">{params}</span>}
       {stats && <span className="flex-shrink-0 text-muted-foreground text-xs">{stats}</span>}
     </div>
   )

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/GlobTool.tsx

@@ -8,7 +8,7 @@ export function GlobTool({
   input,
   output
 }: {
-  input?: GlobToolInputType
+  input: GlobToolInputType
   output?: GlobToolOutputType
 }): NonNullable<CollapseProps['items']>[number] {
   // 如果有输出，计算文件数量
@@ -20,7 +20,7 @@ export function GlobTool({
       <ToolTitle
         icon={<FolderSearch className="h-4 w-4" />}
         label="Glob"
-        params={input?.pattern}
+        params={input.pattern}
         stats={output ? `${lineCount} ${lineCount === 1 ? 'file' : 'files'}` : undefined}
       />
     ),

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/GrepTool.tsx

@@ -8,7 +8,7 @@ export function GrepTool({
   input,
   output
 }: {
-  input?: GrepToolInput
+  input: GrepToolInput
   output?: GrepToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   // 如果有输出，计算结果行数
@@ -22,8 +22,8 @@ export function GrepTool({
         label="Grep"
         params={
           <>
-            {input?.pattern}
-            {input?.output_mode && <span className="ml-1">({input.output_mode})</span>}
+            {input.pattern}
+            {input.output_mode && <span className="ml-1">({input.output_mode})</span>}
           </>
         }
         stats={output ? `${resultLines} ${resultLines === 1 ? 'line' : 'lines'}` : undefined}

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/MultiEditTool.tsx

@@ -9,19 +9,18 @@ import { AgentToolsType } from './types'
 export function MultiEditTool({
   input
 }: {
-  input?: MultiEditToolInput
+  input: MultiEditToolInput
   output?: MultiEditToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
-  const edits = Array.isArray(input?.edits) ? input.edits : []
   return {
     key: AgentToolsType.MultiEdit,
-    label: <ToolTitle icon={<FileText className="h-4 w-4" />} label="MultiEdit" params={input?.file_path} />,
+    label: <ToolTitle icon={<FileText className="h-4 w-4" />} label="MultiEdit" params={input.file_path} />,
     children: (
       <div>
-        {edits.map((edit, index) => (
+        {input.edits.map((edit, index) => (
           <div key={index}>
-            {renderCodeBlock(edit.old_string ?? '', 'old')}
-            {renderCodeBlock(edit.new_string ?? '', 'new')}
+            {renderCodeBlock(edit.old_string, 'old')}
+            {renderCodeBlock(edit.new_string, 'new')}
           </div>
         ))}
       </div>

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/NotebookEditTool.tsx

@@ -11,7 +11,7 @@ export function NotebookEditTool({
   input,
   output
 }: {
-  input?: NotebookEditToolInput
+  input: NotebookEditToolInput
   output?: NotebookEditToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   return {
@@ -20,10 +20,10 @@ export function NotebookEditTool({
       <>
         <ToolTitle icon={<FileText className="h-4 w-4" />} label="NotebookEdit" />
         <Tag className="mt-1" color="blue">
-          {input?.notebook_path}{' '}
+          {input.notebook_path}{' '}
         </Tag>
       </>
     ),
-    children: <ReactMarkdown>{output ?? ''}</ReactMarkdown>
+    children: <ReactMarkdown>{output}</ReactMarkdown>
   }
 }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/ReadTool.tsx

@@ -46,7 +46,7 @@ export function ReadTool({
   input,
   output
 }: {
-  input?: ReadToolInputType
+  input: ReadToolInputType
   output?: ReadToolOutputType
 }): NonNullable<CollapseProps['items']>[number] {
   const outputString = normalizeOutputString(output)
@@ -58,7 +58,7 @@ export function ReadTool({
       <ToolTitle
         icon={<FileText className="h-4 w-4" />}
         label="Read File"
-        params={input?.file_path?.split('/').pop()}
+        params={input.file_path.split('/').pop()}
         stats={stats ? `${stats.lineCount} lines, ${stats.formatSize(stats.fileSize)}` : undefined}
       />
     ),

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/SearchTool.tsx

@@ -8,7 +8,7 @@ export function SearchTool({
   input,
   output
 }: {
-  input?: SearchToolInputType
+  input: SearchToolInputType
   output?: SearchToolOutputType
 }): NonNullable<CollapseProps['items']>[number] {
   // 如果有输出，计算结果数量
@@ -20,13 +20,13 @@ export function SearchTool({
       <ToolTitle
         icon={<Search className="h-4 w-4" />}
         label="Search"
-        params={input ? `"${input}"` : undefined}
+        params={`"${input}"`}
         stats={output ? `${resultCount} ${resultCount === 1 ? 'result' : 'results'}` : undefined}
       />
     ),
     children: (
       <div>
-        {input && <StringInputTool input={input} label="Search Query" />}
+        <StringInputTool input={input} label="Search Query" />
         {output && (
           <div>
             <StringOutputTool output={output} label="Search Results" textColor="text-yellow-600 dark:text-yellow-400" />

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/SkillTool.tsx

@@ -8,12 +8,12 @@ export function SkillTool({
   input,
   output
 }: {
-  input?: SkillToolInput
+  input: SkillToolInput
   output?: SkillToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   return {
     key: 'tool',
-    label: <ToolTitle icon={<PencilRuler className="h-4 w-4" />} label="Skill" params={input?.command} />,
+    label: <ToolTitle icon={<PencilRuler className="h-4 w-4" />} label="Skill" params={input.command} />,
     children: <div>{output}</div>
   }
 }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/TaskTool.tsx

@@ -9,20 +9,19 @@ export function TaskTool({
   input,
   output
 }: {
-  input?: TaskToolInputType
+  input: TaskToolInputType
   output?: TaskToolOutputType
 }): NonNullable<CollapseProps['items']>[number] {
   return {
     key: 'tool',
-    label: <ToolTitle icon={<Bot className="h-4 w-4" />} label="Task" params={input?.description} />,
+    label: <ToolTitle icon={<Bot className="h-4 w-4" />} label="Task" params={input.description} />,
     children: (
       <div>
-        {Array.isArray(output) &&
-          output.map((item) => (
-            <div key={item.type}>
-              <div>{item.type === 'text' ? <Markdown>{item.text}</Markdown> : item.text}</div>
-            </div>
-          ))}
+        {output?.map((item) => (
+          <div key={item.type}>
+            <div>{item.type === 'text' ? <Markdown>{item.text}</Markdown> : item.text}</div>
+          </div>
+        ))}
       </div>
     )
   }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/TodoWriteTool.tsx

@@ -38,10 +38,9 @@ const getStatusConfig = (status: TodoItem['status']) => {
 export function TodoWriteTool({
   input
 }: {
-  input?: TodoWriteToolInputType
+  input: TodoWriteToolInputType
 }): NonNullable<CollapseProps['items']>[number] {
-  const todos = Array.isArray(input?.todos) ? input.todos : []
-  const doneCount = todos.filter((todo) => todo.status === 'completed').length
+  const doneCount = input.todos.filter((todo) => todo.status === 'completed').length
 
   return {
     key: AgentToolsType.TodoWrite,
@@ -50,12 +49,12 @@ export function TodoWriteTool({
         icon={<ListTodo className="h-4 w-4" />}
         label="Todo Write"
         params={`${doneCount} Done`}
-        stats={`${todos.length} ${todos.length === 1 ? 'item' : 'items'}`}
+        stats={`${input.todos.length} ${input.todos.length === 1 ? 'item' : 'items'}`}
       />
     ),
     children: (
       <div className="space-y-3">
-        {todos.map((todo, index) => {
+        {input.todos.map((todo, index) => {
           const statusConfig = getStatusConfig(todo.status)
           return (
             <div key={index}>

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/WebFetchTool.tsx

@@ -8,12 +8,12 @@ export function WebFetchTool({
   input,
   output
 }: {
-  input?: WebFetchToolInput
+  input: WebFetchToolInput
   output?: WebFetchToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   return {
     key: 'tool',
-    label: <ToolTitle icon={<Globe className="h-4 w-4" />} label="Web Fetch" params={input?.url} />,
+    label: <ToolTitle icon={<Globe className="h-4 w-4" />} label="Web Fetch" params={input.url} />,
     children: <div>{output}</div>
   }
 }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/WebSearchTool.tsx

@@ -8,7 +8,7 @@ export function WebSearchTool({
   input,
   output
 }: {
-  input?: WebSearchToolInput
+  input: WebSearchToolInput
   output?: WebSearchToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   // 如果有输出，计算结果数量
@@ -20,7 +20,7 @@ export function WebSearchTool({
       <ToolTitle
         icon={<Globe className="h-4 w-4" />}
         label="Web Search"
-        params={input?.query}
+        params={input.query}
         stats={output ? `${resultCount} ${resultCount === 1 ? 'result' : 'results'}` : undefined}
       />
     ),

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/WriteTool.tsx

@@ -7,12 +7,12 @@ import type { WriteToolInput, WriteToolOutput } from './types'
 export function WriteTool({
   input
 }: {
-  input?: WriteToolInput
+  input: WriteToolInput
   output?: WriteToolOutput
 }): NonNullable<CollapseProps['items']>[number] {
   return {
     key: 'tool',
-    label: <ToolTitle icon={<FileText className="h-4 w-4" />} label="Write" params={input?.file_path} />,
-    children: <div>{input?.content}</div>
+    label: <ToolTitle icon={<FileText className="h-4 w-4" />} label="Write" params={input.file_path} />,
+    children: <div>{input.content}</div>
   }
 }

src/renderer/src/pages/home/Messages/Tools/MessageAgentTools/index.tsx

@@ -1,10 +1,7 @@
 import { loggerService } from '@logger'
-import { useAppSelector } from '@renderer/store'
-import { selectPendingPermission } from '@renderer/store/toolPermissions'
 import type { NormalToolResponse } from '@renderer/types'
 import type { CollapseProps } from 'antd'
-import { Collapse, Spin } from 'antd'
-import { useTranslation } from 'react-i18next'
+import { Collapse } from 'antd'
 
 // 导出所有类型
 export * from './types'
@@ -86,41 +83,17 @@ function ToolContent({ toolName, input, output }: { toolName: AgentToolsType; in
 // 统一的组件渲染入口
 export function MessageAgentTools({ toolResponse }: { toolResponse: NormalToolResponse }) {
   const { arguments: args, response, tool, status } = toolResponse
-  logger.debug('Rendering agent tool response', {
+  logger.info('Rendering agent tool response', {
     tool: tool,
     arguments: args,
-    status,
     response
   })
 
-  const pendingPermission = useAppSelector((state) =>
-    selectPendingPermission(state.toolPermissions, toolResponse.toolCallId)
-  )
-
   if (status === 'pending') {
-    if (pendingPermission) {
-      return <ToolPermissionRequestCard toolResponse={toolResponse} />
-    }
-    return <ToolPendingIndicator toolName={tool?.name} description={tool?.description} />
+    return <ToolPermissionRequestCard toolResponse={toolResponse} />
   }
 
   return (
     <ToolContent toolName={tool.name as AgentToolsType} input={args as ToolInput} output={response as ToolOutput} />
   )
 }
-
-function ToolPendingIndicator({ toolName, description }: { toolName?: string; description?: string }) {
-  const { t } = useTranslation()
-  const label = toolName || t('agent.toolPermission.toolPendingFallback', 'Tool')
-  const detail = description?.trim() || t('agent.toolPermission.executing')
-
-  return (
-    <div className="flex w-full max-w-xl items-center gap-3 rounded-xl border border-default-200 bg-default-100 px-4 py-3 shadow-sm">
-      <Spin size="small" />
-      <div className="flex flex-col gap-1">
-        <span className="font-semibold text-default-700 text-sm">{label}</span>
-        <span className="text-default-500 text-xs">{detail}</span>
-      </div>
-    </div>
-  )
-}

src/renderer/src/pages/home/Tabs/SettingsTab.tsx

@@ -9,7 +9,7 @@ import {
   DEFAULT_TEMPERATURE,
   MAX_CONTEXT_COUNT
 } from '@renderer/config/constant'
-import { isOpenAIModel, isSupportVerbosityModel } from '@renderer/config/models'
+import { isOpenAIModel } from '@renderer/config/models'
 import { UNKNOWN } from '@renderer/config/translate'
 import { useCodeStyle } from '@renderer/context/CodeStyleProvider'
 import { useTheme } from '@renderer/context/ThemeProvider'
@@ -56,7 +56,7 @@ import type { Assistant, AssistantSettings, CodeStyleVarious, MathEngine } from
 import { isGroqSystemProvider, ThemeMode } from '@renderer/types'
 import { modalConfirm } from '@renderer/utils'
 import { getSendMessageShortcutLabel } from '@renderer/utils/input'
-import { isSupportServiceTierProvider, isSupportVerbosityProvider } from '@renderer/utils/provider'
+import { isSupportServiceTierProvider } from '@renderer/utils/provider'
 import { Button, Col, InputNumber, Row, Slider, Switch } from 'antd'
 import { Settings2 } from 'lucide-react'
 import type { FC } from 'react'
@@ -183,10 +183,7 @@ const SettingsTab: FC<Props> = (props) => {
 
   const model = assistant.model || getDefaultModel()
 
-  const showOpenAiSettings =
-    isOpenAIModel(model) ||
-    isSupportServiceTierProvider(provider) ||
-    (isSupportVerbosityModel(model) && isSupportVerbosityProvider(provider))
+  const showOpenAiSettings = isOpenAIModel(model) || isSupportServiceTierProvider(provider)
 
   return (
     <Container className="settings-tab">

src/renderer/src/pages/settings/AboutSettings.tsx

@@ -404,11 +404,11 @@ const UpdateNotesWrapper = styled.div`
   margin: 8px 0;
   background-color: var(--color-bg-2);
   border-radius: 6px;
-  color: var(--color-text-2);
-  font-size: 14px;
 
   p {
     margin: 0;
+    color: var(--color-text-2);
+    font-size: 14px;
   }
 `
 

src/renderer/src/pages/settings/AssistantSettings/AssistantModelSettings.tsx

@@ -135,18 +135,12 @@ const AssistantModelSettings: FC<Props> = ({ assistant, updateAssistant, updateA
           <Input
             value={typeof param.value === 'string' ? param.value : JSON.stringify(param.value, null, 2)}
             onChange={(e) => {
-              // For JSON type parameters, always store the value as a STRING
-              //
-              // Data Flow:
-              // 1. UI stores: { name: "config", value: '{"key":"value"}', type: "json" } ← STRING format
-              // 2. API parses: getCustomParameters() in src/renderer/src/aiCore/utils/reasoning.ts:687-696
-              //                calls JSON.parse() to convert string to object
-              // 3. Request sends: The parsed object is sent to the AI provider
-              //
-              // Previously this code was parsing JSON here and storing
-              // the object directly, which caused getCustomParameters() to fail when trying
-              // to JSON.parse() an already-parsed object.
-              onUpdateCustomParameter(index, 'value', e.target.value)
+              try {
+                const jsonValue = JSON.parse(e.target.value)
+                onUpdateCustomParameter(index, 'value', jsonValue)
+              } catch {
+                onUpdateCustomParameter(index, 'value', e.target.value)
+              }
             }}
           />
         )

src/renderer/src/services/ApiService.ts

@@ -12,7 +12,7 @@ import type { FetchChatCompletionParams } from '@renderer/types'
 import type { Assistant, MCPServer, MCPTool, Model, Provider } from '@renderer/types'
 import type { StreamTextParams } from '@renderer/types/aiCoreTypes'
 import { type Chunk, ChunkType } from '@renderer/types/chunk'
-import type { Message, ResponseError } from '@renderer/types/newMessage'
+import type { Message } from '@renderer/types/newMessage'
 import type { SdkModel } from '@renderer/types/sdk'
 import { removeSpecialCharactersForTopicName, uuid } from '@renderer/utils'
 import { abortCompletion, readyToAbort } from '@renderer/utils/abortController'
@@ -476,7 +476,7 @@ export async function checkApi(provider: Provider, model: Model, timeout = 15000
   } else {
     const abortId = uuid()
     const signal = readyToAbort(abortId)
-    let streamError: ResponseError | undefined
+    let chunkError
     const params: StreamTextParams = {
       system: assistant.prompt,
       prompt: 'hi',
@@ -495,18 +495,19 @@ export async function checkApi(provider: Provider, model: Model, timeout = 15000
       callType: 'check',
       onChunk: (chunk: Chunk) => {
         if (chunk.type === ChunkType.ERROR) {
-          streamError = chunk.error
+          chunkError = chunk.error
         } else {
           abortCompletion(abortId)
         }
       }
     }
 
+    // Try streaming check
     try {
       await ai.completions(model.id, params, config)
     } catch (e) {
-      if (!isAbortError(e) && !isAbortError(streamError)) {
-        throw streamError ?? e
+      if (!isAbortError(e) && !isAbortError(chunkError)) {
+        throw e
       }
     }
   }

src/renderer/src/services/EventService.ts

@@ -28,5 +28,6 @@ export const EVENT_NAMES = {
   RESEND_MESSAGE: 'RESEND_MESSAGE',
   SHOW_MODEL_SELECTOR: 'SHOW_MODEL_SELECTOR',
   EDIT_CODE_BLOCK: 'EDIT_CODE_BLOCK',
-  CHANGE_TOPIC: 'CHANGE_TOPIC'
+  CHANGE_TOPIC: 'CHANGE_TOPIC',
+  CONTINUE_GENERATION: 'CONTINUE_GENERATION'
 }

src/renderer/src/services/messageStreaming/callbacks/baseCallbacks.ts

@@ -194,7 +194,12 @@ export const createBaseCallbacks = (deps: BaseCallbacksDependencies) => {
         }
       }
 
-      const messageUpdates = { status, metrics: response?.metrics, usage: response?.usage }
+      const messageUpdates = {
+        status,
+        metrics: response?.metrics,
+        usage: response?.usage,
+        finishReason: response?.finishReason
+      }
       dispatch(
         newMessagesActions.updateMessage({
           topicId,

src/renderer/src/store/assistants.ts

@@ -216,7 +216,7 @@ const assistantsSlice = createSlice({
         if (agent.id === action.payload.assistantId) {
           for (const key in settings) {
             if (!agent.settings) {
-              agent.settings = { ...DEFAULT_ASSISTANT_SETTINGS }
+              agent.settings = DEFAULT_ASSISTANT_SETTINGS
             }
             agent.settings[key] = settings[key]
           }

src/renderer/src/store/thunk/messageThunk.ts

@@ -14,9 +14,15 @@ import store from '@renderer/store'
 import { updateTopicUpdatedAt } from '@renderer/store/assistants'
 import { type ApiServerConfig, type Assistant, type FileMetadata, type Model, type Topic } from '@renderer/types'
 import type { AgentSessionEntity, GetAgentSessionResponse } from '@renderer/types/agent'
+import type { Chunk } from '@renderer/types/chunk'
 import { ChunkType } from '@renderer/types/chunk'
 import type { FileMessageBlock, ImageMessageBlock, Message, MessageBlock } from '@renderer/types/newMessage'
-import { AssistantMessageStatus, MessageBlockStatus, MessageBlockType } from '@renderer/types/newMessage'
+import {
+  AssistantMessageStatus,
+  MessageBlockStatus,
+  MessageBlockType,
+  UserMessageStatus
+} from '@renderer/types/newMessage'
 import { uuid } from '@renderer/utils'
 import { addAbortController } from '@renderer/utils/abortController'
 import {
@@ -26,6 +32,7 @@ import {
 } from '@renderer/utils/agentSession'
 import {
   createAssistantMessage,
+  createMainTextBlock,
   createTranslationBlock,
   resetAssistantMessage
 } from '@renderer/utils/messageUtils/create'
@@ -1494,6 +1501,230 @@ export const appendAssistantResponseThunk =
     }
   }
 
+/**
+ * Thunk to continue generation from where an assistant message was truncated.
+ * Appends new content to the original truncated message instead of creating new messages.
+ * This avoids issues with APIs that don't support consecutive assistant messages.
+ * @param topicId - The topic ID.
+ * @param truncatedAssistantMessage - The assistant message that was truncated (finishReason: 'length').
+ * @param assistant - The assistant configuration.
+ */
+export const continueGenerationThunk =
+  (topicId: Topic['id'], truncatedAssistantMessage: Message, assistant: Assistant) =>
+  async (dispatch: AppDispatch, getState: () => RootState) => {
+    try {
+      const state = getState()
+
+      // Verify the truncated message exists
+      if (!state.messages.entities[truncatedAssistantMessage.id]) {
+        logger.error(`[continueGenerationThunk] Truncated message ${truncatedAssistantMessage.id} not found.`)
+        return
+      }
+
+      // Get the content of the truncated message to include in the continuation prompt
+      const truncatedContent = getMainTextContent(truncatedAssistantMessage)
+
+      // Create a continuation prompt that asks the AI to continue strictly from where it left off
+      // Use only the last 150 chars to minimize repetition - just enough for context
+      const continuationPrompt = t('message.continue_generation.prompt', {
+        truncatedContent: truncatedContent.slice(-150)
+      })
+
+      // Update the truncated message status to PROCESSING to indicate continuation
+      const messageUpdates = {
+        status: AssistantMessageStatus.PROCESSING,
+        updatedAt: new Date().toISOString()
+      }
+      dispatch(
+        newMessagesActions.updateMessage({
+          topicId,
+          messageId: truncatedAssistantMessage.id,
+          updates: messageUpdates
+        })
+      )
+      dispatch(updateTopicUpdatedAt({ topicId }))
+
+      // Queue the generation with continuation context
+      const queue = getTopicQueue(topicId)
+      const assistantConfig = {
+        ...assistant,
+        model: truncatedAssistantMessage.model || assistant.model
+      }
+      queue.add(async () => {
+        await fetchAndProcessContinuationImpl(
+          dispatch,
+          getState,
+          topicId,
+          assistantConfig,
+          truncatedAssistantMessage,
+          continuationPrompt
+        )
+      })
+    } catch (error) {
+      logger.error(`[continueGenerationThunk] Error continuing generation:`, error as Error)
+    } finally {
+      finishTopicLoading(topicId)
+    }
+  }
+
+/**
+ * Implementation for continuing generation on a truncated message.
+ * Similar to fetchAndProcessAssistantResponseImpl but:
+ * 1. Finds the existing main text block to append content to
+ * 2. Uses a continuation prompt to ask the AI to continue
+ * 3. Wraps the chunk processor to prepend existing content to new text
+ */
+const fetchAndProcessContinuationImpl = async (
+  dispatch: AppDispatch,
+  getState: () => RootState,
+  topicId: string,
+  origAssistant: Assistant,
+  truncatedMessage: Message,
+  continuationPrompt: string
+) => {
+  const topic = origAssistant.topics.find((t) => t.id === topicId)
+  const assistant = topic?.prompt
+    ? { ...origAssistant, prompt: `${origAssistant.prompt}\n${topic.prompt}` }
+    : origAssistant
+  const assistantMsgId = truncatedMessage.id
+  let callbacks: StreamProcessorCallbacks = {}
+
+  // Create a virtual user message with the continuation prompt
+  // We need to temporarily add the block to store so getMainTextContent can read it
+  const virtualUserMessageId = uuid()
+  const virtualTextBlock = createMainTextBlock(virtualUserMessageId, continuationPrompt, {
+    status: MessageBlockStatus.SUCCESS
+  })
+
+  try {
+    dispatch(newMessagesActions.setTopicLoading({ topicId, loading: true }))
+
+    // Find the existing main text block content to prepend
+    const state = getState()
+    const existingMainTextBlockId = truncatedMessage.blocks.find((blockId) => {
+      const block = state.messageBlocks.entities[blockId]
+      return block?.type === MessageBlockType.MAIN_TEXT
+    })
+    const existingContent = existingMainTextBlockId
+      ? (state.messageBlocks.entities[existingMainTextBlockId] as any)?.content || ''
+      : ''
+
+    // Create BlockManager instance
+    const blockManager = new BlockManager({
+      dispatch,
+      getState,
+      saveUpdatedBlockToDB,
+      saveUpdatesToDB,
+      assistantMsgId,
+      topicId,
+      throttledBlockUpdate,
+      cancelThrottledBlockUpdate
+    })
+
+    const allMessagesForTopic = selectMessagesForTopic(getState(), topicId)
+
+    // Find the original user message that triggered this assistant response
+    const userMessageId = truncatedMessage.askId
+    const userMessageIndex = allMessagesForTopic.findIndex((m) => m?.id === userMessageId)
+
+    let messagesForContext: Message[] = []
+    if (userMessageIndex === -1) {
+      logger.error(
+        `[fetchAndProcessContinuationImpl] Triggering user message ${userMessageId} not found. Falling back.`
+      )
+      const assistantMessageIndex = allMessagesForTopic.findIndex((m) => m?.id === assistantMsgId)
+      messagesForContext = (
+        assistantMessageIndex !== -1 ? allMessagesForTopic.slice(0, assistantMessageIndex) : allMessagesForTopic
+      ).filter((m) => m && !m.status?.includes('ing'))
+    } else {
+      // Include messages up to the user message
+      const contextSlice = allMessagesForTopic.slice(0, userMessageIndex + 1)
+      messagesForContext = contextSlice.filter((m) => m && !m.status?.includes('ing'))
+    }
+
+    // Add the truncated assistant message content to context
+    const truncatedAssistantForContext: Message = {
+      ...truncatedMessage,
+      status: AssistantMessageStatus.SUCCESS // Treat as completed for context
+    }
+
+    const virtualContinueMessage: Message = {
+      id: virtualUserMessageId,
+      role: 'user',
+      topicId,
+      assistantId: assistant.id,
+      createdAt: new Date().toISOString(),
+      status: UserMessageStatus.SUCCESS,
+      blocks: [virtualTextBlock.id],
+      model: assistant.model,
+      modelId: assistant.model?.id
+    }
+
+    // Temporarily add the block to store (will be removed in finally block)
+    dispatch(upsertOneBlock(virtualTextBlock))
+
+    // Build the final context: original context + truncated assistant + virtual user message
+    messagesForContext = [...messagesForContext, truncatedAssistantForContext, virtualContinueMessage]
+
+    // Create standard callbacks (no modification needed)
+    callbacks = createCallbacks({
+      blockManager,
+      dispatch,
+      getState,
+      topicId,
+      assistantMsgId,
+      saveUpdatesToDB,
+      assistant
+    })
+    const baseStreamProcessor = createStreamProcessor(callbacks)
+
+    // Wrap the stream processor to prepend existing content to text chunks
+    const wrappedStreamProcessor = (chunk: Chunk) => {
+      if (chunk.type === ChunkType.TEXT_DELTA || chunk.type === ChunkType.TEXT_COMPLETE) {
+        // Prepend existing content to the new text
+        return baseStreamProcessor({
+          ...chunk,
+          text: existingContent + chunk.text
+        })
+      }
+      return baseStreamProcessor(chunk)
+    }
+
+    const abortController = new AbortController()
+    addAbortController(userMessageId!, () => abortController.abort())
+
+    await transformMessagesAndFetch(
+      {
+        messages: messagesForContext,
+        assistant,
+        topicId,
+        options: {
+          signal: abortController.signal,
+          timeout: 30000,
+          headers: defaultAppHeaders()
+        }
+      },
+      wrappedStreamProcessor
+    )
+  } catch (error: any) {
+    logger.error('Error in fetchAndProcessContinuationImpl:', error)
+    endSpan({
+      topicId,
+      error: error,
+      modelName: assistant.model?.name
+    })
+    try {
+      callbacks.onError?.(error)
+    } catch (callbackError) {
+      logger.error('Error in onError callback:', callbackError as Error)
+    }
+  } finally {
+    // Always clean up the temporary virtual block
+    dispatch(removeManyBlocks([virtualTextBlock.id]))
+    dispatch(newMessagesActions.setTopicLoading({ topicId, loading: false }))
+  }
+}
+
 /**
  * Clones messages from a source topic up to a specified index into a *pre-existing* new topic.
  * Generates new unique IDs for all cloned messages and blocks.

src/renderer/src/types/index.ts

@@ -239,7 +239,6 @@ export type ModelType = 'text' | 'vision' | 'embedding' | 'reasoning' | 'functio
 
 export type ModelTag = Exclude<ModelType, 'text'> | 'free'
 
-// "image-generation" is also openai endpoint, but specifically for image generation.
 export type EndpointType = 'openai' | 'openai-response' | 'anthropic' | 'gemini' | 'image-generation' | 'jina-rerank'
 
 export type ModelPricing = {

src/renderer/src/types/newMessage.ts

@@ -1,5 +1,5 @@
 import type { CompletionUsage } from '@cherrystudio/openai/resources'
-import type { ProviderMetadata } from 'ai'
+import type { FinishReason, ProviderMetadata } from 'ai'
 
 import type {
   Assistant,
@@ -221,6 +221,10 @@ export type Message = {
   // raw data
   // TODO: add this providerMetadata to MessageBlock to save raw provider data for each block
   providerMetadata?: ProviderMetadata
+
+  // Finish reason from AI SDK (e.g., 'stop', 'length', 'content-filter', etc.)
+  // Used to show warnings when generation was truncated or filtered
+  finishReason?: FinishReason
 }
 
 export interface Response {
@@ -232,9 +236,9 @@ export interface Response {
   mcpToolResponse?: MCPToolResponse[]
   generateImage?: GenerateImageResponse
   error?: ResponseError
+  finishReason?: FinishReason
 }
 
-// FIXME: Weak type safety. It may be a specific class instance which inherits Error in runtime.
 export type ResponseError = Record<string, any>
 
 export interface MessageInputBaseParams {

src/renderer/src/utils/__tests__/api.test.ts

@@ -7,13 +7,11 @@ import {
   formatApiKeys,
   formatAzureOpenAIApiHost,
   formatVertexApiHost,
-  getTrailingApiVersion,
   hasAPIVersion,
   maskApiKey,
   routeToEndpoint,
   splitApiKeyString,
-  validateApiHost,
-  withoutTrailingApiVersion
+  validateApiHost
 } from '../api'
 
 vi.mock('@renderer/store', () => {
@@ -318,90 +316,4 @@ describe('api', () => {
       )
     })
   })
-
-  describe('getTrailingApiVersion', () => {
-    it('extracts trailing API version from URL', () => {
-      expect(getTrailingApiVersion('https://api.example.com/v1')).toBe('v1')
-      expect(getTrailingApiVersion('https://api.example.com/v2')).toBe('v2')
-    })
-
-    it('extracts trailing API version with alpha/beta suffix', () => {
-      expect(getTrailingApiVersion('https://api.example.com/v2alpha')).toBe('v2alpha')
-      expect(getTrailingApiVersion('https://api.example.com/v3beta')).toBe('v3beta')
-    })
-
-    it('extracts trailing API version with trailing slash', () => {
-      expect(getTrailingApiVersion('https://api.example.com/v1/')).toBe('v1')
-      expect(getTrailingApiVersion('https://api.example.com/v2beta/')).toBe('v2beta')
-    })
-
-    it('returns undefined when API version is in the middle of path', () => {
-      expect(getTrailingApiVersion('https://api.example.com/v1/chat')).toBeUndefined()
-      expect(getTrailingApiVersion('https://api.example.com/v1/completions')).toBeUndefined()
-    })
-
-    it('returns undefined when no trailing version exists', () => {
-      expect(getTrailingApiVersion('https://api.example.com')).toBeUndefined()
-      expect(getTrailingApiVersion('https://api.example.com/api')).toBeUndefined()
-    })
-
-    it('extracts trailing version from complex URLs', () => {
-      expect(getTrailingApiVersion('https://api.example.com/service/v1')).toBe('v1')
-      expect(getTrailingApiVersion('https://gateway.ai.cloudflare.com/v1/xxx/google-ai-studio/v1beta')).toBe('v1beta')
-    })
-
-    it('only extracts the trailing version when multiple versions exist', () => {
-      expect(getTrailingApiVersion('https://api.example.com/v1/service/v2')).toBe('v2')
-      expect(
-        getTrailingApiVersion('https://gateway.ai.cloudflare.com/v1/xxxxxx/google-ai-studio/google-ai-studio/v1beta')
-      ).toBe('v1beta')
-    })
-
-    it('returns undefined for empty string', () => {
-      expect(getTrailingApiVersion('')).toBeUndefined()
-    })
-  })
-
-  describe('withoutTrailingApiVersion', () => {
-    it('removes trailing API version from URL', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/v1')).toBe('https://api.example.com')
-      expect(withoutTrailingApiVersion('https://api.example.com/v2')).toBe('https://api.example.com')
-    })
-
-    it('removes trailing API version with alpha/beta suffix', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/v2alpha')).toBe('https://api.example.com')
-      expect(withoutTrailingApiVersion('https://api.example.com/v3beta')).toBe('https://api.example.com')
-    })
-
-    it('removes trailing API version with trailing slash', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/v1/')).toBe('https://api.example.com')
-      expect(withoutTrailingApiVersion('https://api.example.com/v2beta/')).toBe('https://api.example.com')
-    })
-
-    it('does not remove API version in the middle of path', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/v1/chat')).toBe('https://api.example.com/v1/chat')
-      expect(withoutTrailingApiVersion('https://api.example.com/v1/completions')).toBe(
-        'https://api.example.com/v1/completions'
-      )
-    })
-
-    it('returns URL unchanged when no trailing version exists', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com')).toBe('https://api.example.com')
-      expect(withoutTrailingApiVersion('https://api.example.com/api')).toBe('https://api.example.com/api')
-    })
-
-    it('handles complex URLs with version at the end', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/service/v1')).toBe('https://api.example.com/service')
-    })
-
-    it('handles URLs with multiple versions but only removes the trailing one', () => {
-      expect(withoutTrailingApiVersion('https://api.example.com/v1/service/v2')).toBe(
-        'https://api.example.com/v1/service'
-      )
-    })
-
-    it('returns empty string unchanged', () => {
-      expect(withoutTrailingApiVersion('')).toBe('')
-    })
-  })
 })