LLM_MODEL_PROVIDER2CONFIG#

scikitplot.llm_provider.LLM_MODEL_PROVIDER2CONFIG = {'anthropic': [{'auth': True, 'capabilities': ['chat'], 'context_window': '200k', 'description': 'Optimized for speed and low latency.', 'docs_url': 'https://docs.anthropic.com/claude/docs/models-overview', 'license': 'commercial', 'model_id': 'claude-3-5-haiku-20241022', 'name': 'Claude Haiku 3.5', 'tags': ['fast'], 'version': '2024-10'}, {'auth': True, 'capabilities': ['chat', 'reasoning'], 'context_window': '200k', 'description': 'Most powerful Claude model for advanced reasoning.', 'docs_url': 'https://docs.anthropic.com/claude/docs/models-overview', 'license': 'commercial', 'model_id': 'claude-opus-4-20250514', 'name': 'Claude Opus 4', 'tags': ['premium'], 'version': '2025-05'}, {'auth': True, 'capabilities': ['chat'], 'context_window': '200k', 'description': 'Balanced performance and speed.', 'docs_url': 'https://docs.anthropic.com/claude/docs/models-overview', 'license': 'commercial', 'model_id': 'claude-sonnet-4-20250514', 'name': 'Claude Sonnet 4', 'tags': ['balanced'], 'version': '2025-05'}], 'cohere': [{'auth': True, 'capabilities': ['chat', 'rag'], 'context_window': '128k', 'description': 'Command R is an instruction-following conversational model that performs language tasks at a higher quality, more reliably, and with a longer context than previous models.', 'docs_url': 'https://docs.cohere.com/v2/docs/command-r', 'license': 'open', 'model_id': 'command-r-08-2024', 'name': 'Command R', 'tags': ['rag', 'enhanced'], 'version': '2024-08'}, {'auth': True, 'capabilities': ['chat', 'rag'], 'context_window': '128k', 'description': 'Enhanced version with stronger RAG and reasoning capabilities.', 'docs_url': 'https://docs.cohere.com/v2/docs/command-r-plus', 'license': 'open', 'model_id': 'command-r-plus-08-2024', 'name': 'Command R+', 'tags': ['rag', 'enhanced'], 'version': '2024-08'}, {'auth': True, 'capabilities': ['chat', 'rag'], 'context_window': '', 'description': 'Open weights RAG-tuned model.', 'docs_url': 'https://docs.cohere.com/v2/docs/command-r7b', 'license': 'open', 'model_id': 'command-r7b-12-2024', 'name': 'Command R7B', 'tags': ['rag', 'open-weights'], 'version': '2024-12'}], 'deepseek': [{'auth': False, 'capabilities': ['code', 'chat'], 'context_window': '16k', 'description': 'Code-generation-focused instruction-tuned model.', 'docs_url': 'https://api-docs.deepseek.com', 'license': 'open', 'model_id': 'deepseek-chat', 'name': 'DeepSeek-V3', 'tags': ['developer', 'coding'], 'version': '2023-12'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '16k', 'description': 'Chat model fine-tuned on curated instruction datasets.', 'docs_url': 'https://api-docs.deepseek.com', 'license': 'open', 'model_id': 'deepseek-reasoner', 'name': 'DeepSeek-R1', 'tags': ['deepseek', 'chat'], 'version': '2024-03'}], 'gemini': [{'auth': True, 'capabilities': ['chat', 'text'], 'context_window': '32k', 'description': 'Our best model in terms of price-performance, offering well-rounded capabilities. Gemini 2.5 Flash rate limits are more restricted since it is an experimental / preview model.', 'docs_url': 'https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview', 'license': 'commercial', 'model_id': 'gemini-2.5-flash-preview-05-20', 'name': 'Gemini 2.5 Flash Preview 05-20', 'tags': ['google'], 'version': '2025-05'}, {'auth': True, 'capabilities': ['chat', 'text'], 'context_window': '32k', 'description': 'Gemini 2.0 Flash delivers next-gen features and improved capabilities, including superior speed, native tool use, and a 1M token context window.', 'docs_url': 'https://ai.google.dev/gemini-api/docs/models#gemini-2.0-flash', 'license': 'commercial', 'model_id': 'gemini-2.0-flash', 'name': 'Gemini 2.0 Flash', 'tags': ['google'], 'version': '2025-02'}], 'groq': [{'auth': True, 'capabilities': ['chat'], 'context_window': '8k', 'description': "Meta's LLaMA 3 (8B) on Groq's ultra-fast hardware.", 'docs_url': 'https://groq.com', 'license': 'open', 'model_id': 'llama3-8b-8192', 'name': 'LLaMA 3 8B @ Groq', 'tags': ['llama', 'fast'], 'version': '2024-04'}, {'auth': True, 'capabilities': ['chat', 'code'], 'context_window': '32k', 'description': 'Mixture of Experts model with high performance on Groq.', 'docs_url': 'https://groq.com', 'license': 'open', 'model_id': 'mixtral-8x7b-32768', 'name': 'Mixtral 8x7B @ Groq', 'tags': ['moe', 'performance'], 'version': '2024-03'}], 'huggingface': [{'auth': False, 'capabilities': ['chat'], 'context_window': '', 'description': 'Cohere Labs Command R+ 08-2024 is part of a family of open weight releases from Cohere Labs and Cohere. Our smaller companion model is Cohere Labs Command R', 'docs_url': 'https://huggingface.co/CohereLabs/c4ai-command-r-plus-08-2024', 'license': 'open', 'model_id': 'CohereLabs/c4ai-command-r-plus-08-2024', 'name': 'Command R+ 08-2024', 'tags': ['command', 'instruction-tuned', 'chat'], 'version': '2024-08'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '', 'description': 'RAG with Command R7B is supported through chat templates in Transformers. The model takes a conversation as input (with an optional user-supplied system preamble), along with a list of document snippets.', 'docs_url': 'https://huggingface.co/CohereLabs/c4ai-command-r7b-12-2024', 'license': 'open', 'model_id': 'CohereLabs/c4ai-command-r7b-12-2024', 'name': 'Command R7B 12-2024', 'tags': ['command', 'instruction-tuned', 'chat'], 'version': '2024-12'}, {'auth': False, 'capabilities': ['code', 'chat'], 'context_window': '16k', 'description': 'Code-generation-focused instruction-tuned model.', 'docs_url': 'https://huggingface.co/deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct', 'license': 'open', 'model_id': 'deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct', 'name': 'DeepSeek Coder V2 Lite Instruct', 'tags': ['developer', 'coding'], 'version': '2024-10'}, {'auth': False, 'capabilities': ['code', 'chat'], 'context_window': '16k', 'description': 'Code-generation-focused instruction-tuned model.', 'docs_url': 'https://huggingface.co/deepseek-ai/deepseek-coder-33b-instruct', 'license': 'open', 'model_id': 'deepseek-ai/deepseek-coder-33b-instruct', 'name': 'DeepSeek Coder 33b', 'tags': ['developer', 'coding'], 'version': '2024-08'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '16k', 'description': 'Chat model fine-tuned on curated instruction datasets.', 'docs_url': 'https://huggingface.co/deepseek-ai/deepseek-llm-7b-chat', 'license': 'open', 'model_id': 'deepseek-ai/deepseek-llm-7b-chat', 'name': 'DeepSeek LLM 7B Chat', 'tags': ['deepseek', 'chat'], 'version': '2024-08'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '16k', 'description': 'Fine-tuned open-source chat model.', 'docs_url': 'https://huggingface.co/HuggingFaceH4/zephyr-7b-beta', 'license': 'open', 'model_id': 'HuggingFaceH4/zephyr-7b-beta', 'name': 'Zephyr 7B Beta', 'tags': ['open-source', 'lightweight'], 'version': '2023-10'}, {'auth': False, 'capabilities': ['chat', 'code'], 'context_window': '32k', 'description': 'Instruction-tuned Mistral 7B v0.3 with 32k context window.', 'docs_url': 'https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3', 'license': 'open', 'model_id': 'mistralai/Mistral-7B-Instruct-v0.3', 'name': 'Mistral 7B Instruct v0.3', 'tags': ['mistral', 'updated', '32k'], 'version': '2023-12'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '4k', 'description': 'Small, efficient model from Microsoft, good for edge use.', 'docs_url': 'https://huggingface.co/microsoft/Phi-3-mini-4k-instruct', 'license': 'open', 'model_id': 'microsoft/Phi-3-mini-4k-instruct', 'name': 'Phi-3 Mini', 'tags': ['edge', 'tiny'], 'version': '2024-04'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '4k', 'description': 'Lightweight and performant small model.', 'docs_url': 'https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0', 'license': 'open', 'model_id': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0', 'name': 'TinyLlama Chat', 'tags': ['tiny', 'fast', 'low-resource'], 'version': '2024-01'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '4k', 'description': 'StableVicuna-13B is a Vicuna-13B v0 model fine-tuned using reinforcement learning from human feedback (RLHF) via Proximal Policy Optimization (PPO) on various conversational and instructional datasets.', 'docs_url': 'https://huggingface.co/CarperAI/stable-vicuna-13b-delta', 'license': 'open', 'model_id': 'CarperAI/stable-vicuna-13b-delta', 'name': 'StableVicuna-13B', 'tags': ['vicuna', 'instruction-tuned', 'chat'], 'version': '2023-03'}, {'auth': False, 'capabilities': ['chat'], 'context_window': '4k', 'description': 'Instruction-tuned version of LLaMA 7B, popular open-source chat model.', 'docs_url': 'https://huggingface.co/lmsys/vicuna-7b-v1.5', 'license': 'open', 'model_id': 'lmsys/vicuna-7b-v1.5', 'name': 'Vicuna 7B v1.5', 'tags': ['vicuna', 'instruction-tuned', 'chat'], 'version': '2023-07'}], 'llama': [{'auth': False, 'capabilities': ['chat'], 'context_window': '8k', 'description': 'LLaMA 3 8B chat model, strong open-source performance.', 'docs_url': 'https://huggingface.co/meta-llama/Llama-3-8b-chat-hf', 'license': 'open', 'model_id': 'meta-llama/Llama-3-8b-chat-hf', 'name': 'LLaMA 3 8B Chat', 'tags': ['llama', '8b'], 'version': '2024-04'}, {'auth': False, 'capabilities': ['chat', 'code'], 'context_window': '8k', 'description': '70B SOTA open weights model for enterprise use.', 'docs_url': 'https://huggingface.co/meta-llama/Llama-3-70b-chat-hf', 'license': 'open', 'model_id': 'meta-llama/Llama-3-70b-chat-hf', 'name': 'LLaMA 3 70B Chat', 'tags': ['llama', '70b'], 'version': '2024-04'}], 'mistral': [{'auth': False, 'capabilities': ['chat', 'code'], 'context_window': '32k', 'description': 'Mixture of Experts model from Mistral optimized for instruction following.', 'docs_url': 'https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1', 'license': 'open', 'model_id': 'mistralai/Mixtral-8x7B-Instruct-v0.1', 'name': 'Mixtral 8x7B Instruct', 'tags': ['moe', 'open-source', 'high-performance'], 'version': '2023-12'}], 'openai': [{'auth': True, 'capabilities': ['chat', 'code', 'vision'], 'context_window': '128k', 'description': 'Multimodal GPT-4 with strong reasoning and real-time vision/audio.', 'docs_url': 'https://platform.openai.com/docs/models/gpt-4o', 'license': 'commercial', 'model_id': 'gpt-4o', 'name': 'GPT-4o', 'tags': ['premium', 'multimodal'], 'version': '2024-05'}, {'auth': True, 'capabilities': ['chat', 'code'], 'context_window': '16k', 'description': 'Fast and cost-effective LLM from OpenAI.', 'docs_url': 'https://platform.openai.com/docs/models/gpt-3-5', 'license': 'commercial', 'model_id': 'gpt-3.5-turbo', 'name': 'GPT-3.5 Turbo', 'tags': ['budget'], 'version': '2023-11'}]}#

!! processed by numpydoc !!