add model catalogs

resources/model-catalogs/01-ai/yi-large.yaml

@@ -0,0 +1,47 @@
+id: 01-ai/yi-large
+canonical_slug: 01-ai/yi-large
+hugging_face_id: ''
+name: '01.AI: Yi Large'
+type: chat
+created: 1719273600
+description: |-
+  The Yi Large model was designed by 01.AI with the following usecases in mind: knowledge search, data classification, human-like chat bots, and customer service.
+
+  It stands out for its multilingual proficiency, particularly in Spanish, Chinese, Japanese, German, and French.
+
+  Check out the [launch announcement](https://01-ai.github.io/blog/01.ai-yi-large-llm-launch) to learn more.
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Yi
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000003'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - response_format
+  - structured_outputs
+  - logit_bias
+  - logprobs
+  - top_logprobs
+model_provider: 01-ai

resources/model-catalogs/aetherwiing/mn-starcannon-12b.yaml

@@ -0,0 +1,42 @@
+id: aetherwiing/mn-starcannon-12b
+canonical_slug: aetherwiing/mn-starcannon-12b
+hugging_face_id: aetherwiing/MN-12B-Starcannon-v2
+name: 'Aetherwiing: Starcannon 12B'
+type: chat
+created: 1723507200
+description: |-
+  Starcannon 12B v2 is a creative roleplay and story writing model, based on Mistral Nemo, using [nothingiisreal/mn-celeste-12b](/nothingiisreal/mn-celeste-12b) as a base, with [intervitens/mini-magnum-12b-v1.1](https://huggingface.co/intervitens/mini-magnum-12b-v1.1) merged in using the [TIES](https://arxiv.org/abs/2306.01708) method.
+
+  Although more similar to Magnum overall, the model remains very creative, with a pleasant writing style. It is recommended for people wanting more variety than Magnum, and yet more verbose prose than Celeste.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Mistral
+  instruct_type: chatml
+pricing:
+  prompt: '0.0000008'
+  completion: '0.0000012'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: aetherwiing

resources/model-catalogs/ai21/jamba-1.6-large.yaml

@@ -0,0 +1,38 @@
+id: ai21/jamba-1.6-large
+canonical_slug: ai21/jamba-1.6-large
+hugging_face_id: ai21labs/AI21-Jamba-Large-1.6
+name: 'AI21: Jamba 1.6 Large'
+type: chat
+created: 1741905173
+description: |-
+  AI21 Jamba Large 1.6 is a high-performance hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. Developed by AI21, it excels in extremely long-context handling (256K tokens), demonstrates superior inference efficiency (up to 2.5x faster than comparable models), and supports structured JSON output and tool-use capabilities. It has 94 billion active parameters (398 billion total), optimized quantization support (ExpertsInt8), and multilingual proficiency in languages such as English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew.
+
+  Usage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).
+context_length: 256000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.000002'
+  completion: '0.000008'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+model_provider: ai21

resources/model-catalogs/ai21/jamba-1.6-mini.yaml

@@ -0,0 +1,38 @@
+id: ai21/jamba-1.6-mini
+canonical_slug: ai21/jamba-1.6-mini
+hugging_face_id: ai21labs/AI21-Jamba-Mini-1.6
+name: 'AI21: Jamba Mini 1.6'
+type: chat
+created: 1741905171
+description: |-
+  AI21 Jamba Mini 1.6 is a hybrid foundation model combining State Space Models (Mamba) with Transformer attention mechanisms. With 12 billion active parameters (52 billion total), this model excels in extremely long-context tasks (up to 256K tokens) and achieves superior inference efficiency, outperforming comparable open models on tasks such as retrieval-augmented generation (RAG) and grounded question answering. Jamba Mini 1.6 supports multilingual tasks across English, Spanish, French, Portuguese, Italian, Dutch, German, Arabic, and Hebrew, along with structured JSON output and tool-use capabilities.
+
+  Usage of this model is subject to the [Jamba Open Model License](https://www.ai21.com/licenses/jamba-open-model-license).
+context_length: 256000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000002'
+  completion: '0.0000004'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+model_provider: ai21

resources/model-catalogs/aion-labs/aion-1.0-mini.yaml

@@ -0,0 +1,34 @@
+id: aion-labs/aion-1.0-mini
+canonical_slug: aion-labs/aion-1.0-mini
+hugging_face_id: FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview
+name: 'AionLabs: Aion-1.0-Mini'
+type: chat
+created: 1738697107
+description: Aion-1.0-Mini 32B parameter model is a distilled version of the DeepSeek-R1 model, designed for strong performance in reasoning domains such as mathematics, coding, and logic. It is a modified variant of a FuseAI model that outperforms R1-Distill-Qwen-32B and R1-Distill-Llama-70B, with benchmark results available on its [Hugging Face page](https://huggingface.co/FuseAI/FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview), independently replicated for verification.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000007'
+  completion: '0.0000014'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+model_provider: aion-labs

resources/model-catalogs/aion-labs/aion-1.0.yaml

@@ -0,0 +1,34 @@
+id: aion-labs/aion-1.0
+canonical_slug: aion-labs/aion-1.0
+hugging_face_id: ''
+name: 'AionLabs: Aion-1.0'
+type: chat
+created: 1738697557
+description: Aion-1.0 is a multi-model system designed for high performance across various tasks, including reasoning and coding. It is built on DeepSeek-R1, augmented with additional models and techniques such as Tree of Thoughts (ToT) and Mixture of Experts (MoE). It is Aion Lab's most powerful reasoning model.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.000004'
+  completion: '0.000008'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+model_provider: aion-labs

resources/model-catalogs/aion-labs/aion-rp-llama-3.1-8b.yaml

@@ -0,0 +1,32 @@
+id: aion-labs/aion-rp-llama-3.1-8b
+canonical_slug: aion-labs/aion-rp-llama-3.1-8b
+hugging_face_id: ''
+name: 'AionLabs: Aion-RP 1.0 (8B)'
+type: chat
+created: 1738696718
+description: Aion-RP-Llama-3.1-8B ranks the highest in the character evaluation portion of the RPBench-Auto benchmark, a roleplaying-specific variant of Arena-Hard-Auto, where LLMs evaluate each other’s responses. It is a fine-tuned base model rather than an instruct model, designed to produce more natural and varied writing.
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000002'
+  completion: '0.0000002'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+model_provider: aion-labs

resources/model-catalogs/alfredpros/codellama-7b-instruct-solidity.yaml

@@ -0,0 +1,39 @@
+id: alfredpros/codellama-7b-instruct-solidity
+canonical_slug: alfredpros/codellama-7b-instruct-solidity
+hugging_face_id: AlfredPros/CodeLlama-7b-Instruct-Solidity
+name: 'AlfredPros: CodeLLaMa 7B Instruct Solidity'
+type: chat
+created: 1744641874
+description: A finetuned 7 billion parameters Code LLaMA - Instruct model to generate Solidity smart contract using 4-bit QLoRA finetuning provided by PEFT library.
+context_length: 4096
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: alpaca
+pricing:
+  prompt: '0.0000008'
+  completion: '0.0000012'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: alfredpros

resources/model-catalogs/all-hands/openhands-lm-32b-v0.1.yaml

@@ -0,0 +1,44 @@
+id: all-hands/openhands-lm-32b-v0.1
+canonical_slug: all-hands/openhands-lm-32b-v0.1
+hugging_face_id: all-hands/openhands-lm-32b-v0.1
+name: OpenHands LM 32B V0.1
+type: chat
+created: 1743613013
+description: |-
+  OpenHands LM v0.1 is a 32B open-source coding model fine-tuned from Qwen2.5-Coder-32B-Instruct using reinforcement learning techniques outlined in SWE-Gym. It is optimized for autonomous software development agents and achieves strong performance on SWE-Bench Verified, with a 37.2% resolve rate. The model supports a 128K token context window, making it well-suited for long-horizon code reasoning and large codebase tasks.
+
+  OpenHands LM is designed for local deployment and runs on consumer-grade GPUs such as a single 3090. It enables fully offline agent workflows without dependency on proprietary APIs. This release is intended as a research preview, and future updates aim to improve generalizability, reduce repetition, and offer smaller variants.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000026'
+  completion: '0.0000034'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: all-hands

resources/model-catalogs/alpindale/goliath-120b.yaml

@@ -0,0 +1,48 @@
+id: alpindale/goliath-120b
+canonical_slug: alpindale/goliath-120b
+hugging_face_id: alpindale/goliath-120b
+name: Goliath 120B
+type: chat
+created: 1699574400
+description: |-
+  A large LLM created by combining two fine-tuned Llama 70B models into one 120B model. Combines Xwin and Euryale.
+
+  Credits to
+  - [@chargoddard](https://huggingface.co/chargoddard) for developing the framework used to merge the model - [mergekit](https://github.com/cg123/mergekit).
+  - [@Undi95](https://huggingface.co/Undi95) for helping with the merge ratios.
+
+  #merge
+context_length: 6144
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Llama2
+  instruct_type: airoboros
+pricing:
+  prompt: '0.00001'
+  completion: '0.0000125'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - logit_bias
+  - top_k
+  - min_p
+  - seed
+  - top_a
+model_provider: alpindale

resources/model-catalogs/alpindale/magnum-72b.yaml

@@ -0,0 +1,42 @@
+id: alpindale/magnum-72b
+canonical_slug: alpindale/magnum-72b
+hugging_face_id: alpindale/magnum-72b-v1
+name: Magnum 72B
+type: chat
+created: 1720656000
+description: |-
+  From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the first in a new family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.
+
+  The model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: chatml
+pricing:
+  prompt: '0.000004'
+  completion: '0.000006'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: alpindale

resources/model-catalogs/amazon/nova-lite-v1.yaml

@@ -0,0 +1,39 @@
+id: amazon/nova-lite-v1
+canonical_slug: amazon/nova-lite-v1
+hugging_face_id: ''
+name: 'Amazon: Nova Lite 1.0'
+type: chat
+created: 1733437363
+description: |-
+  Amazon Nova Lite 1.0 is a very low-cost multimodal model from Amazon that focused on fast processing of image, video, and text inputs to generate text output. Amazon Nova Lite can handle real-time customer interactions, document analysis, and visual question-answering tasks with high accuracy.
+
+  With an input context of 300K tokens, it can analyze multiple images or up to 30 minutes of video in a single input.
+context_length: 300000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Nova
+  instruct_type: null
+pricing:
+  prompt: '0.00000006'
+  completion: '0.00000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0.00009'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: amazon

resources/model-catalogs/amazon/nova-micro-v1.yaml

@@ -0,0 +1,35 @@
+id: amazon/nova-micro-v1
+canonical_slug: amazon/nova-micro-v1
+hugging_face_id: ''
+name: 'Amazon: Nova Micro 1.0'
+type: chat
+created: 1733437237
+description: Amazon Nova Micro 1.0 is a text-only model that delivers the lowest latency responses in the Amazon Nova family of models at a very low cost. With a context length of 128K tokens and optimized for speed and cost, Amazon Nova Micro excels at tasks such as text summarization, translation, content classification, interactive chat, and brainstorming. It has  simple mathematical reasoning and coding abilities.
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Nova
+  instruct_type: null
+pricing:
+  prompt: '0.000000035'
+  completion: '0.00000014'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: amazon

resources/model-catalogs/amazon/nova-pro-v1.yaml

@@ -0,0 +1,41 @@
+id: amazon/nova-pro-v1
+canonical_slug: amazon/nova-pro-v1
+hugging_face_id: ''
+name: 'Amazon: Nova Pro 1.0'
+type: chat
+created: 1733436303
+description: |-
+  Amazon Nova Pro 1.0 is a capable multimodal model from Amazon focused on providing a combination of accuracy, speed, and cost for a wide range of tasks. As of December 2024, it achieves state-of-the-art performance on key benchmarks including visual question answering (TextVQA) and video understanding (VATEX).
+
+  Amazon Nova Pro demonstrates strong capabilities in processing both visual and textual information and at analyzing financial documents.
+
+  **NOTE**: Video input is not supported at this time.
+context_length: 300000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Nova
+  instruct_type: null
+pricing:
+  prompt: '0.0000008'
+  completion: '0.0000032'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0.0012'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: amazon

resources/model-catalogs/anthracite-org/magnum-v2-72b.yaml

@@ -0,0 +1,43 @@
+id: anthracite-org/magnum-v2-72b
+canonical_slug: anthracite-org/magnum-v2-72b
+hugging_face_id: anthracite-org/magnum-v2-72b
+name: Magnum v2 72B
+type: chat
+created: 1727654400
+description: |-
+  From the maker of [Goliath](https://openrouter.ai/models/alpindale/goliath-120b), Magnum 72B is the seventh in a family of models designed to achieve the prose quality of the Claude 3 models, notably Opus & Sonnet.
+
+  The model is based on [Qwen2 72B](https://openrouter.ai/models/qwen/qwen-2-72b-instruct) and trained with 55 million tokens of highly curated roleplay (RP) data.
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: chatml
+pricing:
+  prompt: '0.000003'
+  completion: '0.000003'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - logit_bias
+  - top_k
+  - min_p
+  - seed
+model_provider: anthracite-org

resources/model-catalogs/anthracite-org/magnum-v4-72b.yaml

@@ -0,0 +1,44 @@
+id: anthracite-org/magnum-v4-72b
+canonical_slug: anthracite-org/magnum-v4-72b
+hugging_face_id: anthracite-org/magnum-v4-72b
+name: Magnum v4 72B
+type: chat
+created: 1729555200
+description: |-
+  This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus).
+
+  The model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-2.5-72b-instruct).
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: chatml
+pricing:
+  prompt: '0.0000025'
+  completion: '0.000003'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+  - logit_bias
+  - top_a
+model_provider: anthracite-org

resources/model-catalogs/anthropic/claude-2-beta.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2:beta
+canonical_slug: anthropic/claude-2
+hugging_face_id: ''
+name: 'Anthropic: Claude v2 (self-moderated)'
+type: chat
+created: 1700611200
+description: 'Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.'
+context_length: 200000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-2.0-beta.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2.0:beta
+canonical_slug: anthropic/claude-2.0
+hugging_face_id: ''
+name: 'Anthropic: Claude v2.0 (self-moderated)'
+type: chat
+created: 1690502400
+description: Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.
+context_length: 100000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-2.0.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2.0
+canonical_slug: anthropic/claude-2.0
+hugging_face_id: ''
+name: 'Anthropic: Claude v2.0'
+type: chat
+created: 1690502400
+description: Anthropic's flagship model. Superior performance on tasks that require complex reasoning. Supports hundreds of pages of text.
+context_length: 100000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-2.1-beta.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2.1:beta
+canonical_slug: anthropic/claude-2.1
+hugging_face_id: ''
+name: 'Anthropic: Claude v2.1 (self-moderated)'
+type: chat
+created: 1700611200
+description: 'Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.'
+context_length: 200000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-2.1.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2.1
+canonical_slug: anthropic/claude-2.1
+hugging_face_id: ''
+name: 'Anthropic: Claude v2.1'
+type: chat
+created: 1700611200
+description: 'Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.'
+context_length: 200000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-2.yaml

@@ -0,0 +1,34 @@
+id: anthropic/claude-2
+canonical_slug: anthropic/claude-2
+hugging_face_id: ''
+name: 'Anthropic: Claude v2'
+type: chat
+created: 1700611200
+description: 'Claude 2 delivers advancements in key capabilities for enterprises—including an industry-leading 200K token context window, significant reductions in rates of model hallucination, system prompts and a new beta feature: tool use.'
+context_length: 200000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000008'
+  completion: '0.000024'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-haiku-beta.yaml

@@ -0,0 +1,43 @@
+id: anthropic/claude-3-haiku:beta
+canonical_slug: anthropic/claude-3-haiku
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Haiku (self-moderated)'
+type: chat
+created: 1710288000
+description: |-
+  Claude 3 Haiku is Anthropic's fastest and most compact model for
+  near-instant responsiveness. Quick and accurate targeted performance.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.00000025'
+  completion: '0.00000125'
+  input_cache_read: '0.00000003'
+  input_cache_write: '0.0000003'
+  request: '0'
+  image: '0.0004'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-haiku.yaml

@@ -0,0 +1,43 @@
+id: anthropic/claude-3-haiku
+canonical_slug: anthropic/claude-3-haiku
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Haiku'
+type: chat
+created: 1710288000
+description: |-
+  Claude 3 Haiku is Anthropic's fastest and most compact model for
+  near-instant responsiveness. Quick and accurate targeted performance.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-haiku)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.00000025'
+  completion: '0.00000125'
+  input_cache_read: '0.00000003'
+  input_cache_write: '0.0000003'
+  request: '0'
+  image: '0.0004'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-opus-beta.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3-opus:beta
+canonical_slug: anthropic/claude-3-opus
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Opus (self-moderated)'
+type: chat
+created: 1709596800
+description: |-
+  Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000015'
+  completion: '0.000075'
+  input_cache_read: '0.0000015'
+  input_cache_write: '0.00001875'
+  request: '0'
+  image: '0.024'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-opus.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3-opus
+canonical_slug: anthropic/claude-3-opus
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Opus'
+type: chat
+created: 1709596800
+description: |-
+  Claude 3 Opus is Anthropic's most powerful model for highly complex tasks. It boasts top-level performance, intelligence, fluency, and understanding.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000015'
+  completion: '0.000075'
+  input_cache_read: '0.0000015'
+  input_cache_write: '0.00001875'
+  request: '0'
+  image: '0.024'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-sonnet-beta.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3-sonnet:beta
+canonical_slug: anthropic/claude-3-sonnet
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Sonnet (self-moderated)'
+type: chat
+created: 1709596800
+description: |-
+  Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3-sonnet.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3-sonnet
+canonical_slug: anthropic/claude-3-sonnet
+hugging_face_id: ''
+name: 'Anthropic: Claude 3 Sonnet'
+type: chat
+created: 1709596800
+description: |-
+  Claude 3 Sonnet is an ideal balance of intelligence and speed for enterprise workloads. Maximum utility at a lower price, dependable, balanced for scaled deployments.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/claude-3-family)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-haiku-20241022-beta.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3.5-haiku-20241022:beta
+canonical_slug: anthropic/claude-3-5-haiku-20241022
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Haiku (2024-10-22) (self-moderated)'
+type: chat
+created: 1730678400
+description: |-
+  Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.
+
+  It does not support image inputs.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.0000008'
+  completion: '0.000004'
+  input_cache_read: '0.00000008'
+  input_cache_write: '0.000001'
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-haiku-20241022.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3.5-haiku-20241022
+canonical_slug: anthropic/claude-3-5-haiku-20241022
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Haiku (2024-10-22)'
+type: chat
+created: 1730678400
+description: |-
+  Claude 3.5 Haiku features enhancements across all skill sets including coding, tool use, and reasoning. As the fastest model in the Anthropic lineup, it offers rapid response times suitable for applications that require high interactivity and low latency, such as user-facing chatbots and on-the-fly code completions. It also excels in specialized tasks like data extraction and real-time content moderation, making it a versatile tool for a broad range of industries.
+
+  It does not support image inputs.
+
+  See the launch announcement and benchmark results [here](https://www.anthropic.com/news/3-5-models-and-computer-use)
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.0000008'
+  completion: '0.000004'
+  input_cache_read: '0.00000008'
+  input_cache_write: '0.000001'
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-haiku-beta.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3.5-haiku:beta
+canonical_slug: anthropic/claude-3-5-haiku
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Haiku (self-moderated)'
+type: chat
+created: 1730678400
+description: |-
+  Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.
+
+  This makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.
+
+  This model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.0000008'
+  completion: '0.000004'
+  input_cache_read: '0.00000008'
+  input_cache_write: '0.000001'
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-haiku.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-3.5-haiku
+canonical_slug: anthropic/claude-3-5-haiku
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Haiku'
+type: chat
+created: 1730678400
+description: |-
+  Claude 3.5 Haiku features offers enhanced capabilities in speed, coding accuracy, and tool use. Engineered to excel in real-time applications, it delivers quick response times that are essential for dynamic tasks such as chat interactions and immediate coding suggestions.
+
+  This makes it highly suitable for environments that demand both speed and precision, such as software development, customer service bots, and data management systems.
+
+  This model is currently pointing to [Claude 3.5 Haiku (2024-10-22)](/anthropic/claude-3-5-haiku-20241022).
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.0000008'
+  completion: '0.000004'
+  input_cache_read: '0.00000008'
+  input_cache_write: '0.000001'
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-sonnet-20240620-beta.yaml

@@ -0,0 +1,47 @@
+id: anthropic/claude-3.5-sonnet-20240620:beta
+canonical_slug: anthropic/claude-3.5-sonnet-20240620
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Sonnet (2024-06-20) (self-moderated)'
+type: chat
+created: 1718841600
+description: |-
+  Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:
+
+  - Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting
+  - Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights
+  - Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone
+  - Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)
+
+  For the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-sonnet-20240620.yaml

@@ -0,0 +1,47 @@
+id: anthropic/claude-3.5-sonnet-20240620
+canonical_slug: anthropic/claude-3.5-sonnet-20240620
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Sonnet (2024-06-20)'
+type: chat
+created: 1718841600
+description: |-
+  Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:
+
+  - Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting
+  - Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights
+  - Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone
+  - Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)
+
+  For the latest version (2024-10-23), check out [Claude 3.5 Sonnet](/anthropic/claude-3.5-sonnet).
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-sonnet-beta.yaml

@@ -0,0 +1,45 @@
+id: anthropic/claude-3.5-sonnet:beta
+canonical_slug: anthropic/claude-3.5-sonnet
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Sonnet (self-moderated)'
+type: chat
+created: 1729555200
+description: |-
+  New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:
+
+  - Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding
+  - Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights
+  - Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone
+  - Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.5-sonnet.yaml

@@ -0,0 +1,45 @@
+id: anthropic/claude-3.5-sonnet
+canonical_slug: anthropic/claude-3.5-sonnet
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.5 Sonnet'
+type: chat
+created: 1729555200
+description: |-
+  New Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:
+
+  - Coding: Scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding
+  - Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights
+  - Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone
+  - Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)
+
+  #multimodal
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - top_k
+  - stop
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.7-sonnet-beta.yaml

@@ -0,0 +1,37 @@
+id: anthropic/claude-3.7-sonnet:beta
+canonical_slug: anthropic/claude-3-7-sonnet-20250219
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.7 Sonnet (self-moderated)'
+type: chat
+created: 1740422110
+description: "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)"
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.7-sonnet-thinking.yaml

@@ -0,0 +1,37 @@
+id: anthropic/claude-3.7-sonnet:thinking
+canonical_slug: anthropic/claude-3-7-sonnet-20250219
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.7 Sonnet (thinking)'
+type: chat
+created: 1740422110
+description: "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)"
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-3.7-sonnet.yaml

@@ -0,0 +1,39 @@
+id: anthropic/claude-3.7-sonnet
+canonical_slug: anthropic/claude-3-7-sonnet-20250219
+hugging_face_id: ''
+name: 'Anthropic: Claude 3.7 Sonnet'
+type: chat
+created: 1740422110
+description: "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. \n\nClaude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks.\n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)"
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-opus-4.yaml

@@ -0,0 +1,39 @@
+id: anthropic/claude-opus-4
+canonical_slug: anthropic/claude-4-opus-20250522
+hugging_face_id: ''
+name: 'Anthropic: Claude Opus 4'
+type: chat
+created: 1747931245
+description: "Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows. It sets new benchmarks in software engineering, achieving leading results on SWE-bench (72.5%) and Terminal-bench (43.2%). Opus 4 supports extended, agentic workflows, handling thousands of task steps continuously for hours without degradation. \n\nRead more at the [blog post here](https://www.anthropic.com/news/claude-4)"
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000015'
+  completion: '0.000075'
+  input_cache_read: '0.0000015'
+  input_cache_write: '0.00001875'
+  request: '0'
+  image: '0.024'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+model_provider: anthropic

resources/model-catalogs/anthropic/claude-sonnet-4.yaml

@@ -0,0 +1,42 @@
+id: anthropic/claude-sonnet-4
+canonical_slug: anthropic/claude-4-sonnet-20250522
+hugging_face_id: ''
+name: 'Anthropic: Claude Sonnet 4'
+type: chat
+created: 1747930371
+description: |-
+  Claude Sonnet 4 significantly enhances the capabilities of its predecessor, Sonnet 3.7, excelling in both coding and reasoning tasks with improved precision and controllability. Achieving state-of-the-art performance on SWE-bench (72.7%), Sonnet 4 balances capability and computational efficiency, making it suitable for a broad range of applications from routine coding tasks to complex software development projects. Key enhancements include improved autonomous codebase navigation, reduced error rates in agent-driven workflows, and increased reliability in following intricate instructions. Sonnet 4 is optimized for practical everyday use, providing advanced reasoning capabilities while maintaining efficiency and responsiveness in diverse internal and external scenarios.
+
+  Read more at the [blog post here](https://www.anthropic.com/news/claude-4)
+context_length: 200000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Claude
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: '0.0000003'
+  input_cache_write: '0.00000375'
+  request: '0'
+  image: '0.0048'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+model_provider: anthropic

resources/model-catalogs/arcee-ai/arcee-blitz.yaml

@@ -0,0 +1,40 @@
+id: arcee-ai/arcee-blitz
+canonical_slug: arcee-ai/arcee-blitz
+hugging_face_id: arcee-ai/arcee-blitz
+name: 'Arcee AI: Arcee Blitz'
+type: chat
+created: 1746470100
+description: 'Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for "everyday" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price. '
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.00000045'
+  completion: '0.00000075'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/caller-large.yaml

@@ -0,0 +1,42 @@
+id: arcee-ai/caller-large
+canonical_slug: arcee-ai/caller-large
+hugging_face_id: ''
+name: 'Arcee AI: Caller Large'
+type: chat
+created: 1746487869
+description: 'Caller Large is Arcee''s specialist "function‑calling" SLM built to orchestrate external tools and APIs. Instead of maximizing next‑token accuracy, training focuses on structured JSON outputs, parameter extraction and multi‑step tool chains, making Caller a natural choice for retrieval‑augmented generation, robotic process automation or data‑pull chatbots. It incorporates a routing head that decides when (and how) to invoke a tool versus answering directly, reducing hallucinated calls. The model is already the backbone of Arcee Conductor''s auto‑tool mode, where it parses user intent, emits clean function signatures and hands control back once the tool response is ready. Developers thus gain an OpenAI‑style function‑calling UX without handing requests to a frontier‑scale model. '
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.00000055'
+  completion: '0.00000085'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/coder-large.yaml

@@ -0,0 +1,40 @@
+id: arcee-ai/coder-large
+canonical_slug: arcee-ai/coder-large
+hugging_face_id: ''
+name: 'Arcee AI: Coder Large'
+type: chat
+created: 1746478663
+description: 'Coder‑Large is a 32 B‑parameter offspring of Qwen 2.5‑Instruct that has been further trained on permissively‑licensed GitHub, CodeSearchNet and synthetic bug‑fix corpora. It supports a 32k context window, enabling multi‑file refactoring or long diff review in a single call, and understands 30‑plus programming languages with special attention to TypeScript, Go and Terraform. Internal benchmarks show 5–8 pt gains over CodeLlama‑34 B‑Python on HumanEval and competitive BugFix scores thanks to a reinforcement pass that rewards compilable output. The model emits structured explanations alongside code blocks by default, making it suitable for educational tooling as well as production copilot scenarios. Cost‑wise, Together AI prices it well below proprietary incumbents, so teams can scale interactive coding without runaway spend. '
+context_length: 32768
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000005'
+  completion: '0.0000008'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/maestro-reasoning.yaml

@@ -0,0 +1,40 @@
+id: arcee-ai/maestro-reasoning
+canonical_slug: arcee-ai/maestro-reasoning
+hugging_face_id: ''
+name: 'Arcee AI: Maestro Reasoning'
+type: chat
+created: 1746481269
+description: 'Maestro Reasoning is Arcee''s flagship analysis model: a 32 B‑parameter derivative of Qwen 2.5‑32 B tuned with DPO and chain‑of‑thought RL for step‑by‑step logic. Compared to the earlier 7 B preview, the production 32 B release widens the context window to 128 k tokens and doubles pass‑rate on MATH and GSM‑8K, while also lifting code completion accuracy. Its instruction style encourages structured "thought → answer" traces that can be parsed or hidden according to user preference. That transparency pairs well with audit‑focused industries like finance or healthcare where seeing the reasoning path matters. In Arcee Conductor, Maestro is automatically selected for complex, multi‑constraint queries that smaller SLMs bounce. '
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000009'
+  completion: '0.0000033'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/spotlight.yaml

@@ -0,0 +1,41 @@
+id: arcee-ai/spotlight
+canonical_slug: arcee-ai/spotlight
+hugging_face_id: ''
+name: 'Arcee AI: Spotlight'
+type: chat
+created: 1746481552
+description: 'Spotlight is a 7‑billion‑parameter vision‑language model derived from Qwen 2.5‑VL and fine‑tuned by Arcee AI for tight image‑text grounding tasks. It offers a 32 k‑token context window, enabling rich multimodal conversations that combine lengthy documents with one or more images. Training emphasized fast inference on consumer GPUs while retaining strong captioning, visual‐question‑answering, and diagram‑analysis accuracy. As a result, Spotlight slots neatly into agent workflows where screenshots, charts or UI mock‑ups need to be interpreted on the fly. Early benchmarks show it matching or out‑scoring larger VLMs such as LLaVA‑1.6 13 B on popular VQA and POPE alignment tests. '
+context_length: 131072
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.00000018'
+  completion: '0.00000018'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/virtuoso-large.yaml

@@ -0,0 +1,42 @@
+id: arcee-ai/virtuoso-large
+canonical_slug: arcee-ai/virtuoso-large
+hugging_face_id: ''
+name: 'Arcee AI: Virtuoso Large'
+type: chat
+created: 1746478885
+description: Virtuoso‑Large is Arcee's top‑tier general‑purpose LLM at 72 B parameters, tuned to tackle cross‑domain reasoning, creative writing and enterprise QA. Unlike many 70 B peers, it retains the 128 k context inherited from Qwen 2.5, letting it ingest books, codebases or financial filings wholesale. Training blended DeepSeek R1 distillation, multi‑epoch supervised fine‑tuning and a final DPO/RLHF alignment stage, yielding strong performance on BIG‑Bench‑Hard, GSM‑8K and long‑context Needle‑In‑Haystack tests. Enterprises use Virtuoso‑Large as the "fallback" brain in Conductor pipelines when other SLMs flag low confidence. Despite its size, aggressive KV‑cache optimizations keep first‑token latency in the low‑second range on 8× H100 nodes, making it a practical production‑grade powerhouse.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.00000075'
+  completion: '0.0000012'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/arcee-ai/virtuoso-medium-v2.yaml

@@ -0,0 +1,42 @@
+id: arcee-ai/virtuoso-medium-v2
+canonical_slug: arcee-ai/virtuoso-medium-v2
+hugging_face_id: arcee-ai/Virtuoso-Medium-v2
+name: 'Arcee AI: Virtuoso Medium V2'
+type: chat
+created: 1746478434
+description: 'Virtuoso‑Medium‑v2 is a 32 B model distilled from DeepSeek‑v3 logits and merged back onto a Qwen 2.5 backbone, yielding a sharper, more factual successor to the original Virtuoso Medium. The team harvested ~1.1 B logit tokens and applied "fusion‑merging" plus DPO alignment, which pushed scores past Arcee‑Nova 2024 and many 40 B‑plus peers on MMLU‑Pro, MATH and HumanEval. With a 128 k context and aggressive quantization options (from BF16 down to 4‑bit GGUF), it balances capability with deployability on single‑GPU nodes. Typical use cases include enterprise chat assistants, technical writing aids and medium‑complexity code drafting where Virtuoso‑Large would be overkill. '
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000005'
+  completion: '0.0000008'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: arcee-ai

resources/model-catalogs/bytedance/doubao-embedding-large-text-240715.yaml

@@ -0,0 +1,24 @@
+id: bytedance/doubao-embedding-text-240715
+canonical_slug: bytedance/doubao-embedding-text-240715
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding Text (240715)'
+description: |-
+  Doubao Embedding Large 是字节跳动语义向量化模型的最新升级版，模型以豆包语言模型为基座，具备强大的语言理解能力；主要面向向量检索的使用场景，支持中、英双语。
+context_length: 4000
+dimensions:
+  - 512
+  - 1024
+  - 2048
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.7'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-embedding-large-text-240915.yaml

@@ -0,0 +1,25 @@
+id: bytedance/doubao-embedding-large-text-240915
+canonical_slug: bytedance/doubao-embedding-large-text-240915
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding Large Text (240915)'
+description: |-
+  Doubao Embedding Large 是字节跳动语义向量化模型的最新升级版，模型以豆包语言模型为基座，具备强大的语言理解能力；主要面向向量检索的使用场景，支持中、英双语。
+context_length: 4000
+dimensions:
+  - 512
+  - 1024
+  - 2048
+  - 4096
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.7'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-embedding-text-240515.yaml

@@ -0,0 +1,24 @@
+id: bytedance/doubao-embedding-text-240715
+canonical_slug: bytedance/doubao-embedding-text-240715
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding'
+description: |-
+  由字节跳动研发的语义向量化模型，主要面向向量检索的使用场景，支持中、英双语，最长 4K 上下文长度。向量维度 2048 维，支持 512、1024 降维使用。
+context_length: 4000
+dimensions:
+  - 512
+  - 1024
+  - 2048
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.5'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-embedding-text-240715.yaml

@@ -0,0 +1,25 @@
+id: bytedance/doubao-embedding-text-240715
+canonical_slug: bytedance/doubao-embedding-text-240715
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding'
+description: |-
+  由字节跳动研发的语义向量化模型，主要面向向量检索的使用场景，支持中、英双语，最长 4K 上下文长度。向量维度 2048 维，支持 512、1024 降维使用。
+context_length: 4000
+dimensions:
+  - 512
+  - 1024
+  - 2048
+  - 2560
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.5'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-embedding-vision-241215.yaml

@@ -0,0 +1,24 @@
+id: bytedance/doubao-embedding-vision-241215
+canonical_slug: bytedance/doubao-embedding-vision-241215
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding Vision'
+description: |-
+  Doubao-embedding-vision，全新升级图文多模态向量化模型，主要面向图文多模向量检索的使用场景，支持图片输入及中、英双语文本输入，最长 8K 上下文长度。
+context_length: 8000
+dimensions:
+  - 3072
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.7'
+  prompt_image: '1.8'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-embedding-vision-250328.yaml

@@ -0,0 +1,25 @@
+id: bytedance/doubao-embedding-vision-250328
+canonical_slug: bytedance/doubao-embedding-vision-250328
+type: embedding
+hugging_face_id: null
+name: 'ByteDance: Doubao Embedding Vision'
+description: |-
+  Doubao-embedding-vision，全新升级图文多模态向量化模型，主要面向图文多模向量检索的使用场景，支持图片输入及中、英双语文本输入，最长 8K 上下文长度。
+context_length: 8000
+dimensions:
+  - 1024
+  - 2048
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Doubao
+pricing:
+  prompt: '0.7'
+  prompt_image: '1.8'
+  unit: 1000000
+  currency: CNY
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-seed-1.6-flash.yaml

@@ -0,0 +1,41 @@
+id: bytedance/doubao-seed-1.6-flash
+canonical_slug: bytedance/doubao-seed-1.6-flash
+type: chat
+hugging_face_id: ''
+name: 'ByteDance: Doubao Seed 1.6 Flash'
+created: 1738402289
+description: 有极致推理速度的多模态深度思考模型；同时支持文本和视觉理解。文本理解能力超过上一代 Lite 系列模型，视觉理解比肩友商 Pro 系列模型。
+context_length: 256000
+architecture:
+  modality: text+image+vedio->text
+  input_modalities:
+    - text
+    - image
+    - video
+  output_modalities:
+    - text
+  tokenizer: Doubao
+  instruct_type: null
+pricing:
+  prompt: '0.15'
+  completion: '1.5'
+  input_cache_read: '0.03'
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  currency: CNY
+  unit: 1000000
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+  - structured_outputs
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-seed-1.6-thinking.yaml

@@ -0,0 +1,41 @@
+id: bytedance/doubao-seed-1.6-thinking
+canonical_slug: bytedance/doubao-seed-1.6-thinking
+type: chat
+hugging_face_id: ''
+name: 'ByteDance: Doubao Seed 1.6 Thinking'
+created: 1738402289
+description: 在思考能力上进行了大幅强化， 对比 doubao 1.5 代深度理解模型，在编程、数学、逻辑推理等基础能力上进一步提升， 支持视觉理解。
+context_length: 256000
+architecture:
+  modality: text+image+vedio->text
+  input_modalities:
+    - text
+    - image
+    - video
+  output_modalities:
+    - text
+  tokenizer: Doubao
+  instruct_type: null
+pricing:
+  prompt: '0.8'
+  completion: '8.0'
+  input_cache_read: '0.16'
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  currency: CNY
+  unit: 1000000
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+  - structured_outputs
+model_provider: bytedance

resources/model-catalogs/bytedance/doubao-seed-1.6.yaml

@@ -0,0 +1,41 @@
+id: bytedance/doubao-seed-1.6
+canonical_slug: bytedance/doubao-seed-1.6
+type: chat
+hugging_face_id: ''
+name: 'ByteDance: Doubao Seed 1.6'
+created: 1738402289
+description: 全新多模态深度思考模型，同时支持 thinking、non-thinking、auto三种思考模式。其中 non-thinking 模型对比 doubao-1.5-pro-32k-250115 模型大幅提升。
+context_length: 256000
+architecture:
+  modality: text+image+vedio->text
+  input_modalities:
+    - text
+    - image
+    - video
+  output_modalities:
+    - text
+  tokenizer: Doubao
+  instruct_type: null
+pricing:
+  prompt: '0.8'
+  completion: '8.0'
+  input_cache_read: '0.16'
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  currency: CNY
+  unit: 1000000
+supported_parameters:
+  - max_tokens
+  - temperature
+  - stop
+  - reasoning
+  - include_reasoning
+  - tools
+  - tool_choice
+  - top_p
+  - top_k
+  - structured_outputs
+model_provider: bytedance

resources/model-catalogs/cognitivecomputations/dolphin-mixtral-8x22b.yaml

@@ -0,0 +1,47 @@
+id: cognitivecomputations/dolphin-mixtral-8x22b
+canonical_slug: cognitivecomputations/dolphin-mixtral-8x22b
+hugging_face_id: cognitivecomputations/dolphin-2.9.2-mixtral-8x22b
+name: "Dolphin 2.9.2 Mixtral 8x22B \U0001F42C"
+type: chat
+created: 1717804800
+description: |-
+  Dolphin 2.9 is designed for instruction following, conversational, and coding. This model is a finetune of [Mixtral 8x22B Instruct](/models/mistralai/mixtral-8x22b-instruct). It features a 64k context length and was fine-tuned with a 16k sequence length using ChatML templates.
+
+  This model is a successor to [Dolphin Mixtral 8x7B](/models/cognitivecomputations/dolphin-mixtral-8x7b).
+
+  The model is uncensored and is stripped of alignment and bias. It requires an external alignment layer for ethical use. Users are cautioned to use this highly compliant model responsibly, as detailed in a blog post about uncensored models at [erichartford.com/uncensored-models](https://erichartford.com/uncensored-models).
+
+  #moe #uncensored
+context_length: 16000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Mistral
+  instruct_type: chatml
+pricing:
+  prompt: '0.0000009'
+  completion: '0.0000009'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - top_k
+  - min_p
+  - repetition_penalty
+  - logit_bias
+model_provider: cognitivecomputations

resources/model-catalogs/cohere/command-a.yaml

@@ -0,0 +1,41 @@
+id: cohere/command-a
+canonical_slug: cohere/command-a-03-2025
+hugging_face_id: CohereForAI/c4ai-command-a-03-2025
+name: 'Cohere: Command A'
+type: chat
+created: 1741894342
+description: |-
+  Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.
+  Compared to other leading proprietary and open-weights models Command A delivers maximum performance with minimum hardware costs, excelling on business-critical agentic and multilingual tasks.
+context_length: 256000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: null
+pricing:
+  prompt: '0.0000025'
+  completion: '0.00001'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r-03-2024.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r-03-2024
+canonical_slug: cohere/command-r-03-2024
+hugging_face_id: ''
+name: 'Cohere: Command R (03-2024)'
+type: chat
+created: 1709341200
+description: |-
+  Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.
+
+  Read the launch post [here](https://txt.cohere.com/command-r/).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.0000005'
+  completion: '0.0000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r-08-2024.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r-08-2024
+canonical_slug: cohere/command-r-08-2024
+hugging_face_id: ''
+name: 'Cohere: Command R (08-2024)'
+type: chat
+created: 1724976000
+description: |-
+  command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual retrieval-augmented generation (RAG) and tool use. More broadly, it is better at math, code and reasoning and is competitive with the previous version of the larger Command R+ model.
+
+  Read the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.00000015'
+  completion: '0.0000006'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r-plus-04-2024.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r-plus-04-2024
+canonical_slug: cohere/command-r-plus-04-2024
+hugging_face_id: ''
+name: 'Cohere: Command R+ (04-2024)'
+type: chat
+created: 1712016000
+description: |-
+  Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).
+
+  It offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r-plus-08-2024.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r-plus-08-2024
+canonical_slug: cohere/command-r-plus-08-2024
+hugging_face_id: ''
+name: 'Cohere: Command R+ (08-2024)'
+type: chat
+created: 1724976000
+description: |-
+  command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version, while keeping the hardware footprint the same.
+
+  Read the launch post [here](https://docs.cohere.com/changelog/command-gets-refreshed).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.0000025'
+  completion: '0.00001'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r-plus.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r-plus
+canonical_slug: cohere/command-r-plus
+hugging_face_id: ''
+name: 'Cohere: Command R+'
+type: chat
+created: 1712188800
+description: |-
+  Command R+ is a new, 104B-parameter LLM from Cohere. It's useful for roleplay, general consumer usecases, and Retrieval Augmented Generation (RAG).
+
+  It offers multilingual support for ten key languages to facilitate global business operations. See benchmarks and the launch post [here](https://txt.cohere.com/command-r-plus-microsoft-azure/).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.000003'
+  completion: '0.000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r.yaml

@@ -0,0 +1,45 @@
+id: cohere/command-r
+canonical_slug: cohere/command-r
+hugging_face_id: ''
+name: 'Cohere: Command R'
+type: chat
+created: 1710374400
+description: |-
+  Command-R is a 35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents.
+
+  Read the launch post [here](https://txt.cohere.com/command-r/).
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.0000005'
+  completion: '0.0000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command-r7b-12-2024.yaml

@@ -0,0 +1,42 @@
+id: cohere/command-r7b-12-2024
+canonical_slug: cohere/command-r7b-12-2024
+hugging_face_id: ''
+name: 'Cohere: Command R7B (12-2024)'
+type: chat
+created: 1734158152
+description: |-
+  Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG, tool use, agents, and similar tasks requiring complex reasoning and multiple steps.
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.0000000375'
+  completion: '0.00000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/cohere/command.yaml

@@ -0,0 +1,42 @@
+id: cohere/command
+canonical_slug: cohere/command
+hugging_face_id: ''
+name: 'Cohere: Command'
+type: chat
+created: 1710374400
+description: |-
+  Command is an instruction-following conversational model that performs language tasks with high quality, more reliably and with a longer context than our base generative models.
+
+  Use of this model is subject to Cohere's [Usage Policy](https://docs.cohere.com/docs/usage-policy) and [SaaS Agreement](https://cohere.com/saas-agreement).
+context_length: 4096
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Cohere
+  instruct_type: null
+pricing:
+  prompt: '0.000001'
+  completion: '0.000002'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: cohere

resources/model-catalogs/deepseek/deepseek-chat-v3-0324.yaml

@@ -0,0 +1,49 @@
+id: deepseek/deepseek-chat-v3-0324
+canonical_slug: deepseek/deepseek-chat-v3-0324
+hugging_face_id: deepseek-ai/DeepSeek-V3-0324
+name: 'DeepSeek: DeepSeek V3 0324'
+type: chat
+created: 1742824755
+description: |-
+  DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team.
+
+  It succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well on a variety of tasks.
+context_length: 163840
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: DeepSeek
+  instruct_type: null
+pricing:
+  prompt: '0.0000003'
+  completion: '0.00000088'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - logprobs
+  - top_logprobs
+  - seed
+  - min_p
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-chat.yaml

@@ -0,0 +1,49 @@
+id: deepseek/deepseek-chat
+canonical_slug: deepseek/deepseek-chat-v3
+hugging_face_id: deepseek-ai/DeepSeek-V3
+name: 'DeepSeek: DeepSeek V3'
+type: chat
+created: 1735241320
+description: |-
+  DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.
+
+  For model details, please visit [the DeepSeek-V3 repo](https://github.com/deepseek-ai/DeepSeek-V3) for more information, or see the [launch announcement](https://api-docs.deepseek.com/news/news1226).
+context_length: 163840
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: DeepSeek
+  instruct_type: null
+pricing:
+  prompt: '0.00000038'
+  completion: '0.00000089'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - logprobs
+  - top_logprobs
+  - seed
+  - min_p
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-prover-v2.yaml

@@ -0,0 +1,41 @@
+id: deepseek/deepseek-prover-v2
+canonical_slug: deepseek/deepseek-prover-v2
+hugging_face_id: deepseek-ai/DeepSeek-Prover-V2-671B
+name: 'DeepSeek: DeepSeek Prover V2'
+type: chat
+created: 1746013094
+description: DeepSeek Prover V2 is a 671B parameter model, speculated to be geared towards logic and mathematics. Likely an upgrade from [DeepSeek-Prover-V1.5](https://huggingface.co/deepseek-ai/DeepSeek-Prover-V1.5-RL) Not much is known about the model yet, as DeepSeek released it on Hugging Face without an announcement or description.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: DeepSeek
+  instruct_type: null
+pricing:
+  prompt: '0.0000005'
+  completion: '0.00000218'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - top_k
+  - min_p
+  - repetition_penalty
+  - logit_bias
+  - response_format
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-0528-qwen3-8b.yaml

@@ -0,0 +1,45 @@
+id: deepseek/deepseek-r1-0528-qwen3-8b
+canonical_slug: deepseek/deepseek-r1-0528-qwen3-8b
+hugging_face_id: deepseek-ai/deepseek-r1-0528-qwen3-8b
+name: 'DeepSeek: Deepseek R1 0528 Qwen3 8B'
+type: chat
+created: 1748538543
+description: |-
+  DeepSeek-R1-0528 is a lightly upgraded release of DeepSeek R1 that taps more compute and smarter post-training tricks, pushing its reasoning and inference to the brink of flagship models like O3 and Gemini 2.5 Pro.
+  It now tops math, programming, and logic leaderboards, showcasing a step-change in depth-of-thought.
+  The distilled variant, DeepSeek-R1-0528-Qwen3-8B, transfers this chain-of-thought into an 8 B-parameter form, beating standard Qwen3 8B by +10 pp and tying the 235 B “thinking” giant on AIME 2024.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000005'
+  completion: '0.0000001'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - presence_penalty
+  - frequency_penalty
+  - repetition_penalty
+  - top_k
+  - stop
+  - seed
+  - min_p
+  - logit_bias
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-0528.yaml

@@ -0,0 +1,51 @@
+id: deepseek/deepseek-r1-0528
+canonical_slug: deepseek/deepseek-r1-0528
+hugging_face_id: deepseek-ai/DeepSeek-R1-0528
+name: 'DeepSeek: R1 0528'
+type: chat
+created: 1748455170
+description: |-
+  May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.
+
+  Fully open-source model.
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: DeepSeek
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.0000005'
+  completion: '0.00000215'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+  - logprobs
+  - top_logprobs
+  - tools
+  - tool_choice
+  - seed
+  - structured_outputs
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-llama-70b.yaml

@@ -0,0 +1,55 @@
+id: deepseek/deepseek-r1-distill-llama-70b
+canonical_slug: deepseek/deepseek-r1-distill-llama-70b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Llama-70B
+name: 'DeepSeek: R1 Distill Llama 70B'
+type: chat
+created: 1737663169
+description: |-
+  DeepSeek R1 Distill Llama 70B is a distilled large language model based on [Llama-3.3-70B-Instruct](/meta-llama/llama-3.3-70b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:
+
+  - AIME 2024 pass@1: 70.0
+  - MATH-500 pass@1: 94.5
+  - CodeForces Rating: 1633
+
+  The model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Llama3
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.0000001'
+  completion: '0.0000004'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - seed
+  - top_k
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - logit_bias
+  - logprobs
+  - top_logprobs
+  - min_p
+  - repetition_penalty
+  - tools
+  - tool_choice
+  - response_format
+  - structured_outputs
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-llama-8b.yaml

@@ -0,0 +1,42 @@
+id: deepseek/deepseek-r1-distill-llama-8b
+canonical_slug: deepseek/deepseek-r1-distill-llama-8b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Llama-8B
+name: 'DeepSeek: R1 Distill Llama 8B'
+type: chat
+created: 1738937718
+description: "DeepSeek R1 Distill Llama 8B is a distilled large language model based on [Llama-3.1-8B-Instruct](/meta-llama/llama-3.1-8b-instruct), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). The model combines advanced distillation techniques to achieve high performance across multiple benchmarks, including:\n\n- AIME 2024 pass@1: 50.4\n- MATH-500 pass@1: 89.1\n- CodeForces Rating: 1205\n\nThe model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.\n\nHugging Face: \n- [Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) \n- [DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B)   |"
+context_length: 32000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Llama3
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000004'
+  completion: '0.00000004'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - top_k
+  - min_p
+  - repetition_penalty
+  - logit_bias
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-qwen-1.5b.yaml

@@ -0,0 +1,51 @@
+id: deepseek/deepseek-r1-distill-qwen-1.5b
+canonical_slug: deepseek/deepseek-r1-distill-qwen-1.5b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
+name: 'DeepSeek: R1 Distill Qwen 1.5B'
+type: chat
+created: 1738328067
+description: |-
+  DeepSeek R1 Distill Qwen 1.5B is a distilled large language model based on  [Qwen 2.5 Math 1.5B](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It's a very small and efficient model which outperforms [GPT 4o 0513](/openai/gpt-4o-2024-05-13) on Math Benchmarks.
+
+  Other benchmark results include:
+
+  - AIME 2024 pass@1: 28.9
+  - AIME 2024 cons@64: 52.7
+  - MATH-500 pass@1: 83.9
+
+  The model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000018'
+  completion: '0.00000018'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-qwen-14b.yaml

@@ -0,0 +1,52 @@
+id: deepseek/deepseek-r1-distill-qwen-14b
+canonical_slug: deepseek/deepseek-r1-distill-qwen-14b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
+name: 'DeepSeek: R1 Distill Qwen 14B'
+type: chat
+created: 1738193940
+description: |-
+  DeepSeek R1 Distill Qwen 14B is a distilled large language model based on [Qwen 2.5 14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI's o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.
+
+  Other benchmark results include:
+
+  - AIME 2024 pass@1: 69.7
+  - MATH-500 pass@1: 93.9
+  - CodeForces Rating: 1481
+
+  The model leverages fine-tuning from DeepSeek R1's outputs, enabling competitive performance comparable to larger frontier models.
+context_length: 64000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000015'
+  completion: '0.00000015'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - seed
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - min_p
+  - repetition_penalty
+  - logit_bias
+  - response_format
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-qwen-32b.yaml

@@ -0,0 +1,43 @@
+id: deepseek/deepseek-r1-distill-qwen-32b
+canonical_slug: deepseek/deepseek-r1-distill-qwen-32b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
+name: 'DeepSeek: R1 Distill Qwen 32B'
+type: chat
+created: 1738194830
+description: 'DeepSeek R1 Distill Qwen 32B is a distilled large language model based on [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B), using outputs from [DeepSeek R1](/deepseek/deepseek-r1). It outperforms OpenAI''s o1-mini across various benchmarks, achieving new state-of-the-art results for dense models.\n\nOther benchmark results include:\n\n- AIME 2024 pass@1: 72.6\n- MATH-500 pass@1: 94.3\n- CodeForces Rating: 1691\n\nThe model leverages fine-tuning from DeepSeek R1''s outputs, enabling competitive performance comparable to larger frontier models.'
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000012'
+  completion: '0.00000018'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - seed
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - min_p
+  - repetition_penalty
+  - logit_bias
+  - response_format
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1-distill-qwen-7b.yaml

@@ -0,0 +1,35 @@
+id: deepseek/deepseek-r1-distill-qwen-7b
+canonical_slug: deepseek/deepseek-r1-distill-qwen-7b
+hugging_face_id: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
+name: 'DeepSeek: R1 Distill Qwen 7B'
+type: chat
+created: 1748628237
+description: DeepSeek-R1-Distill-Qwen-7B is a 7 billion parameter dense language model distilled from DeepSeek-R1, leveraging reinforcement learning-enhanced reasoning data generated by DeepSeek's larger models. The distillation process transfers advanced reasoning, math, and code capabilities into a smaller, more efficient model architecture based on Qwen2.5-Math-7B. This model demonstrates strong performance across mathematical benchmarks (92.8% pass@1 on MATH-500), coding tasks (Codeforces rating 1189), and general reasoning (49.1% pass@1 on GPQA Diamond), achieving competitive accuracy relative to larger models while maintaining smaller inference costs.
+context_length: 131072
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.0000001'
+  completion: '0.0000002'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - seed
+model_provider: deepseek

resources/model-catalogs/deepseek/deepseek-r1.yaml

@@ -0,0 +1,53 @@
+id: deepseek/deepseek-r1
+canonical_slug: deepseek/deepseek-r1
+hugging_face_id: deepseek-ai/DeepSeek-R1
+name: 'DeepSeek: R1'
+type: chat
+created: 1737381095
+description: |-
+  DeepSeek R1 is here: Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass.
+
+  Fully open-source model & [technical report](https://api-docs.deepseek.com/news/news250120).
+
+  MIT licensed: Distill & commercialize freely!
+context_length: 128000
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: DeepSeek
+  instruct_type: deepseek-r1
+pricing:
+  prompt: '0.00000045'
+  completion: '0.00000215'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - top_k
+  - min_p
+  - logit_bias
+  - top_logprobs
+  - response_format
+  - structured_outputs
+  - logprobs
+  - repetition_penalty
+  - tools
+  - tool_choice
+model_provider: deepseek

resources/model-catalogs/eleutherai/llemma_7b.yaml

@@ -0,0 +1,39 @@
+id: eleutherai/llemma_7b
+canonical_slug: eleutherai/llemma_7b
+hugging_face_id: EleutherAI/llemma_7b
+name: 'EleutherAI: Llemma 7b'
+type: chat
+created: 1744643225
+description: Llemma 7B is a language model for mathematics. It was initialized with Code Llama 7B weights, and trained on the Proof-Pile-2 for 200B tokens. Llemma models are particularly strong at chain-of-thought mathematical reasoning and using computational tools for mathematics, such as Python and formal theorem provers.
+context_length: 4096
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Other
+  instruct_type: code-llama
+pricing:
+  prompt: '0.0000008'
+  completion: '0.0000012'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: eleutherai

resources/model-catalogs/eva-unit-01/eva-llama-3.33-70b.yaml

@@ -0,0 +1,44 @@
+id: eva-unit-01/eva-llama-3.33-70b
+canonical_slug: eva-unit-01/eva-llama-3.33-70b
+hugging_face_id: EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1
+name: EVA Llama 3.33 70B
+type: chat
+created: 1734377303
+description: |
+  EVA Llama 3.33 70b is a roleplay and storywriting specialist model. It is a full-parameter finetune of [Llama-3.3-70B-Instruct](https://openrouter.ai/meta-llama/llama-3.3-70b-instruct) on mixture of synthetic and natural data.
+
+  It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and "flavor" of the resulting model
+
+  This model was built with Llama by Meta.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Llama3
+  instruct_type: llama3
+pricing:
+  prompt: '0.000004'
+  completion: '0.000006'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: eva-unit-01

resources/model-catalogs/eva-unit-01/eva-qwen-2.5-32b.yaml

@@ -0,0 +1,42 @@
+id: eva-unit-01/eva-qwen-2.5-32b
+canonical_slug: eva-unit-01/eva-qwen-2.5-32b
+hugging_face_id: EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2
+name: EVA Qwen2.5 32B
+type: chat
+created: 1731104847
+description: |-
+  EVA Qwen2.5 32B is a roleplaying/storywriting specialist model. It's a full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data.
+
+  It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and "flavor" of the resulting model.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: chatml
+pricing:
+  prompt: '0.0000026'
+  completion: '0.0000034'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: eva-unit-01

resources/model-catalogs/eva-unit-01/eva-qwen-2.5-72b.yaml

@@ -0,0 +1,42 @@
+id: eva-unit-01/eva-qwen-2.5-72b
+canonical_slug: eva-unit-01/eva-qwen-2.5-72b
+hugging_face_id: EVA-UNIT-01/EVA-Qwen2.5-72B-v0.1
+name: EVA Qwen2.5 72B
+type: chat
+created: 1732210606
+description: |-
+  EVA Qwen2.5 72B is a roleplay and storywriting specialist model. It's a full-parameter finetune of Qwen2.5-72B on mixture of synthetic and natural data.
+
+  It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and "flavor" of the resulting model.
+context_length: 16384
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Qwen
+  instruct_type: chatml
+pricing:
+  prompt: '0.000004'
+  completion: '0.000006'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - repetition_penalty
+  - top_k
+  - min_p
+  - seed
+model_provider: eva-unit-01

resources/model-catalogs/google/gemini-2.0-flash-001.yaml

@@ -0,0 +1,42 @@
+id: google/gemini-2.0-flash-001
+canonical_slug: google/gemini-2.0-flash-001
+hugging_face_id: ''
+name: 'Google: Gemini 2.0 Flash'
+type: chat
+created: 1738769413
+description: Gemini Flash 2.0 offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5). It introduces notable enhancements in multimodal understanding, coding capabilities, complex instruction following, and function calling. These advancements come together to deliver more seamless and robust agentic experiences.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.0000001'
+  completion: '0.0000004'
+  input_cache_read: '0.000000025'
+  input_cache_write: '0.0000001833'
+  request: '0'
+  image: '0.0000258'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.0-flash-lite-001.yaml

@@ -0,0 +1,42 @@
+id: google/gemini-2.0-flash-lite-001
+canonical_slug: google/gemini-2.0-flash-lite-001
+hugging_face_id: ''
+name: 'Google: Gemini 2.0 Flash Lite'
+type: chat
+created: 1740506212
+description: Gemini 2.0 Flash Lite offers a significantly faster time to first token (TTFT) compared to [Gemini Flash 1.5](/google/gemini-flash-1.5), while maintaining quality on par with larger models like [Gemini Pro 1.5](/google/gemini-pro-1.5), all at extremely economical token prices.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.000000075'
+  completion: '0.0000003'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash-lite-preview-06-17.yaml

@@ -0,0 +1,44 @@
+id: google/gemini-2.5-flash-lite-preview-06-17
+canonical_slug: google/gemini-2.5-flash-lite-preview-06-17
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash Lite Preview 06-17'
+type: chat
+created: 1750173831
+description: 'Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, "thinking" (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the [Reasoning API parameter](https://openrouter.ai/docs/use-cases/reasoning-tokens) to selectively trade off cost for intelligence. '
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - file
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.0000001'
+  completion: '0.0000004'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash-preview-05-20-thinking.yaml

@@ -0,0 +1,44 @@
+id: google/gemini-2.5-flash-preview-05-20:thinking
+canonical_slug: google/gemini-2.5-flash-preview-05-20
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash Preview 05-20 (thinking)'
+type: chat
+created: 1747761924
+description: "Gemini 2.5 Flash May 20th Checkpoint is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning)."
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000015'
+  completion: '0.0000035'
+  input_cache_read: '0.0000000375'
+  input_cache_write: '0.0000002333'
+  request: '0'
+  image: '0.0006192'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash-preview-05-20.yaml

@@ -0,0 +1,44 @@
+id: google/gemini-2.5-flash-preview-05-20
+canonical_slug: google/gemini-2.5-flash-preview-05-20
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash Preview 05-20'
+type: chat
+created: 1747761924
+description: "Gemini 2.5 Flash May 20th Checkpoint is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning)."
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000015'
+  completion: '0.0000006'
+  input_cache_read: '0.0000000375'
+  input_cache_write: '0.0000002333'
+  request: '0'
+  image: '0.0006192'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash-preview-thinking.yaml

@@ -0,0 +1,39 @@
+id: google/gemini-2.5-flash-preview:thinking
+canonical_slug: google/gemini-2.5-flash-preview-04-17
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash Preview 04-17 (thinking)'
+type: chat
+created: 1744914667
+description: "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning)."
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000015'
+  completion: '0.0000035'
+  input_cache_read: '0.0000000375'
+  input_cache_write: '0.0000002333'
+  request: '0'
+  image: '0.0006192'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - tools
+  - tool_choice
+  - stop
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash-preview.yaml

@@ -0,0 +1,39 @@
+id: google/gemini-2.5-flash-preview
+canonical_slug: google/gemini-2.5-flash-preview-04-17
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash Preview 04-17'
+type: chat
+created: 1744914667
+description: "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nNote: This model is available in two variants: thinking and non-thinking. The output pricing varies significantly depending on whether the thinking capability is active. If you select the standard variant (without the \":thinking\" suffix), the model will explicitly avoid generating thinking tokens. \n\nTo utilize the thinking capability and receive thinking tokens, you must choose the \":thinking\" variant, which will then incur the higher thinking-output pricing. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning)."
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - image
+    - text
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000015'
+  completion: '0.0000006'
+  input_cache_read: '0.0000000375'
+  input_cache_write: '0.0000002333'
+  request: '0'
+  image: '0.0006192'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - tools
+  - tool_choice
+  - stop
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.5-flash.yaml

@@ -0,0 +1,44 @@
+id: google/gemini-2.5-flash
+canonical_slug: google/gemini-2.5-flash
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Flash'
+type: chat
+created: 1750172488
+description: "Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks. It includes built-in \"thinking\" capabilities, enabling it to provide responses with greater accuracy and nuanced context handling. \n\nAdditionally, Gemini 2.5 Flash is configurable through the \"max tokens for reasoning\" parameter, as described in the documentation (https://openrouter.ai/docs/use-cases/reasoning-tokens#max-tokens-for-reasoning)."
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - file
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.0000003'
+  completion: '0.0000025'
+  input_cache_read: '0.000000075'
+  input_cache_write: '0.0000003833'
+  request: '0'
+  image: '0.001238'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-2.5-pro-exp-03-25.yaml

@@ -0,0 +1,43 @@
+id: google/gemini-2.5-pro-exp-03-25
+canonical_slug: google/gemini-2.5-pro-exp-03-25
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Pro Experimental'
+type: chat
+created: 1742922099
+description: |-
+  This model has been deprecated by Google in favor of the (paid Preview model)[google/gemini-2.5-pro-preview]
+   
+  Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0'
+  completion: '0'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - tools
+  - tool_choice
+  - stop
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.5-pro-preview-05-06.yaml

@@ -0,0 +1,40 @@
+id: google/gemini-2.5-pro-preview-05-06
+canonical_slug: google/gemini-2.5-pro-preview-03-25
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Pro Preview 05-06'
+type: chat
+created: 1746578513
+description: Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+    - file
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000125'
+  completion: '0.00001'
+  input_cache_read: '0.00000031'
+  input_cache_write: '0.000001625'
+  request: '0'
+  image: '0.00516'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - tools
+  - tool_choice
+  - stop
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-2.5-pro-preview.yaml

@@ -0,0 +1,45 @@
+id: google/gemini-2.5-pro-preview
+canonical_slug: google/gemini-2.5-pro-preview-06-05
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Pro Preview 06-05'
+type: chat
+created: 1749137257
+description: |
+  Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - file
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000125'
+  completion: '0.00001'
+  input_cache_read: '0.00000031'
+  input_cache_write: '0.000001625'
+  request: '0'
+  image: '0.00516'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-2.5-pro.yaml

@@ -0,0 +1,44 @@
+id: google/gemini-2.5-pro
+canonical_slug: google/gemini-2.5-pro
+hugging_face_id: ''
+name: 'Google: Gemini 2.5 Pro'
+type: chat
+created: 1750169544
+description: Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking” capabilities, enabling it to reason through responses with enhanced accuracy and nuanced context handling. Gemini 2.5 Pro achieves top-tier performance on multiple benchmarks, including first-place positioning on the LMArena leaderboard, reflecting superior human-preference alignment and complex problem-solving abilities.
+context_length: 1048576
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - file
+    - image
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000125'
+  completion: '0.00001'
+  input_cache_read: '0.00000031'
+  input_cache_write: '0.000001625'
+  request: '0'
+  image: '0.00516'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - tools
+  - tool_choice
+  - max_tokens
+  - temperature
+  - top_p
+  - reasoning
+  - include_reasoning
+  - structured_outputs
+  - response_format
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - seed
+model_provider: google

resources/model-catalogs/google/gemini-flash-1.5-8b.yaml

@@ -0,0 +1,46 @@
+id: google/gemini-flash-1.5-8b
+canonical_slug: google/gemini-flash-1.5-8b
+hugging_face_id: ''
+name: 'Google: Gemini 1.5 Flash 8B'
+type: chat
+created: 1727913600
+description: |-
+  Gemini Flash 1.5 8B is optimized for speed and efficiency, offering enhanced performance in small prompt tasks like chat, transcription, and translation. With reduced latency, it is highly effective for real-time and large-scale operations. This model focuses on cost-effective solutions while maintaining high-quality results.
+
+  [Click here to learn more about this model](https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/).
+
+  Usage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).
+context_length: 1000000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.0000000375'
+  completion: '0.00000015'
+  input_cache_read: '0.00000001'
+  input_cache_write: '0.0000000583'
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - tools
+  - tool_choice
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-flash-1.5.yaml

@@ -0,0 +1,48 @@
+id: google/gemini-flash-1.5
+canonical_slug: google/gemini-flash-1.5
+hugging_face_id: ''
+name: 'Google: Gemini 1.5 Flash '
+type: chat
+created: 1715644800
+description: |-
+  Gemini 1.5 Flash is a foundation model that performs well at a variety of multimodal tasks such as visual understanding, classification, summarization, and creating content from image, audio and video. It's adept at processing visual and text inputs such as photographs, documents, infographics, and screenshots.
+
+  Gemini 1.5 Flash is designed for high-volume, high-frequency tasks where cost and latency matter. On most common tasks, Flash achieves comparable quality to other Gemini Pro models at a significantly reduced cost. Flash is well-suited for applications like chat assistants and on-demand content generation where speed and scale matter.
+
+  Usage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).
+
+  #multimodal
+context_length: 1000000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.000000075'
+  completion: '0.0000003'
+  input_cache_read: '0.00000001875'
+  input_cache_write: '0.0000001583'
+  request: '0'
+  image: '0.00004'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - tools
+  - tool_choice
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemini-pro-1.5.yaml

@@ -0,0 +1,57 @@
+id: google/gemini-pro-1.5
+canonical_slug: google/gemini-pro-1.5
+hugging_face_id: ''
+name: 'Google: Gemini 1.5 Pro'
+type: chat
+created: 1712620800
+description: |-
+  Google's latest multimodal model, supports image and video[0] in text or chat prompts.
+
+  Optimized for language tasks including:
+
+  - Code generation
+  - Text generation
+  - Text editing
+  - Problem solving
+  - Recommendations
+  - Information extraction
+  - Data extraction or generation
+  - AI agents
+
+  Usage of Gemini is subject to Google's [Gemini Terms of Use](https://ai.google.dev/terms).
+
+  * [0]: Video input is not available through OpenRouter at this time.
+context_length: 2000000
+architecture:
+  modality: text+image->text
+  input_modalities:
+    - text
+    - image
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: null
+pricing:
+  prompt: '0.00000125'
+  completion: '0.000005'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0.0006575'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - tools
+  - tool_choice
+  - seed
+  - response_format
+  - structured_outputs
+model_provider: google

resources/model-catalogs/google/gemma-2-27b-it.yaml

@@ -0,0 +1,45 @@
+id: google/gemma-2-27b-it
+canonical_slug: google/gemma-2-27b-it
+hugging_face_id: google/gemma-2-27b-it
+name: 'Google: Gemma 2 27B'
+type: chat
+created: 1720828800
+description: |-
+  Gemma 2 27B by Google is an open model built from the same research and technology used to create the [Gemini models](/models?q=gemini).
+
+  Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning.
+
+  See the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).
+context_length: 8192
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: gemma
+pricing:
+  prompt: '0.0000008'
+  completion: '0.0000008'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - top_k
+  - repetition_penalty
+  - logit_bias
+  - min_p
+  - response_format
+model_provider: google

resources/model-catalogs/google/gemma-2-9b-it.yaml

@@ -0,0 +1,45 @@
+id: google/gemma-2-9b-it
+canonical_slug: google/gemma-2-9b-it
+hugging_face_id: google/gemma-2-9b-it
+name: 'Google: Gemma 2 9B'
+type: chat
+created: 1719532800
+description: |-
+  Gemma 2 9B by Google is an advanced, open-source language model that sets a new standard for efficiency and performance in its size class.
+
+  Designed for a wide variety of tasks, it empowers developers and researchers to build innovative applications, while maintaining accessibility, safety, and cost-effectiveness.
+
+  See the [launch announcement](https://blog.google/technology/developers/google-gemma-2/) for more details. Usage of Gemma is subject to Google's [Gemma Terms of Use](https://ai.google.dev/gemma/terms).
+context_length: 8192
+architecture:
+  modality: text->text
+  input_modalities:
+    - text
+  output_modalities:
+    - text
+  tokenizer: Gemini
+  instruct_type: gemma
+pricing:
+  prompt: '0.0000002'
+  completion: '0.0000002'
+  input_cache_read: ''
+  input_cache_write: ''
+  request: '0'
+  image: '0'
+  web_search: '0'
+  internal_reasoning: '0'
+  unit: 1
+  currency: USD
+supported_parameters:
+  - max_tokens
+  - temperature
+  - top_p
+  - stop
+  - frequency_penalty
+  - presence_penalty
+  - response_format
+  - top_logprobs
+  - logprobs
+  - logit_bias
+  - seed
+model_provider: google