{"coreweave":{"id":"coreweave","name":"CoreWeave","npm":"@ai-sdk/openai-compatible","env":["WANDB_API_KEY"],"doc":"https://docs.wandb.ai/inference","models":{"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"OpenAI: gpt-oss-120b","description":"Efficient Mixture-of-Experts model designed for high-reasoning, agentic and general-purpose use cases.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2025-08-05","last_updated":"2025-08-05","open_weights":true,"cost":{"input":0.03,"output":0.17,"cache_read":0.03},"limit":{"context":131072,"input":131072,"output":131072},"modalities":{"input":["text"],"output":["text"]}},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"OpenAI: gpt-oss-20b","description":"Lower latency Mixture-of-Experts model trained on OpenAI's Harmony response format with reasoning capabilities.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2025-08-05","last_updated":"2025-08-05","open_weights":true,"cost":{"input":0.03,"output":0.13,"cache_read":0.03},"limit":{"context":131072,"input":131072,"output":131072},"modalities":{"input":["text"],"output":["text"]}},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Google: Gemma 4 31B","description":"Gemma 4 31B Dense is designed for advanced reasoning, agentic workflows, and longer context and is natively trained on 140+ languages.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-04-02","last_updated":"2026-04-02","open_weights":true,"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Google: Gemma 4 26B A4B","description":"Gemma 4 26B A4B is a multimodal MoE model with LoRA support and function calling for agentic workflows.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-04-02","last_updated":"2026-04-02","open_weights":true,"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"IBM: Granite 4.2 8B","description":"Granite 4.2 8B is an instruct model capable of enhanced tool calling, instruction following, and chat capabilities.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-08-25","last_updated":"2026-08-25","open_weights":true,"cost":{"input":0.1,"output":0.15,"cache_read":0.05},"limit":{"context":131072,"input":131072,"output":131072},"modalities":{"input":["text"],"output":["text"]}},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax: MiniMax M3","description":"MiniMax M3 is a multimodal MoE model with 23B active parameters optimized for coding and agentic workflows.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-06-12","last_updated":"2026-06-12","open_weights":true,"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"Z.ai: GLM 5.3 Flash","description":"GLM-5.3-Flash is a natively multimodal model with 320B total parameters and 18B active parameters.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-08-26","last_updated":"2026-08-26","open_weights":true,"cost":{"input":0.15,"output":0.5,"cache_read":0.05},"limit":{"context":1048576,"input":1048576,"output":1048576},"modalities":{"input":["text","image"],"output":["text"]}},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"Z.AI: GLM 5.2","description":"GLM-5.2 is a Mixture-of-Experts language model featuring 40 billion activated parameters and a total of 744 billion parameters.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-06-16","last_updated":"2026-06-16","open_weights":true,"cost":{"input":0.76,"output":2.42,"cache_read":0.14},"limit":{"context":1048576,"input":1048576,"output":1048576},"modalities":{"input":["text"],"output":["text"]}},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B","name":"NVIDIA: Nemotron 3.5 Lightning","description":"Nemotron 3.5 Lightning is an MoE model built for fast, reliable agentic tasks across use cases such as financial services, cybersecurity, telecom, and retail.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-08-11","last_updated":"2026-08-11","open_weights":true,"cost":{"input":0.07,"output":0.2,"cache_read":0.04},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text"],"output":["text"]}},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B","name":"NVIDIA: Nemotron 3 Ultra","description":"Nemotron 3 Ultra is a powerful MoE model designed for long-running agents across coding, deep research, and enterprise automation.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-06-04","last_updated":"2026-06-04","open_weights":true,"cost":{"input":0.5,"output":2.15,"cache_read":0.1},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text"],"output":["text"]}},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen: Qwen3.8 27B","description":"Qwen3.8-27B is a dense multimodal model suited for coding, research, vision, and long-running agent tasks.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-08-14","last_updated":"2026-08-14","open_weights":true,"cost":{"input":0.4,"output":3.0,"cache_read":0.15},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen: Qwen3.6 35B A3B","description":"Qwen3.6-35B-A3B is an MoE multimodal model with 262K context optimized for agentic coding workflows.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-04-15","last_updated":"2026-04-15","open_weights":true,"cost":{"input":0.25,"output":1.25,"cache_read":0.25},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Moonshot AI: Kimi K2.7 Code","description":"Kimi K2.7 Code is a 1T-parameter MoE model with 32B active parameters purpose-built for long-horizon agentic coding and software engineering.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-06-12","last_updated":"2026-06-12","open_weights":true,"cost":{"input":0.71,"output":3.5,"cache_read":0.15},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Moonshot AI: Kimi K2.6","description":"Kimi K2.6 is a multimodal Mixture-of-Experts language model featuring 32 billion activated parameters and a total of 1 trillion parameters.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-04-20","last_updated":"2026-04-20","open_weights":true,"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text","image"],"output":["text"]}},"meta-llama/Llama-3.1-8B-Instruct":{"id":"meta-llama/Llama-3.1-8B-Instruct","name":"Meta: Llama 3.1 8B","description":"Efficient conversational model optimized for responsive multilingual chatbot interactions.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2024-07-01","last_updated":"2024-07-01","open_weights":true,"cost":{"input":0.22,"output":0.22,"cache_read":0.22},"limit":{"context":131072,"input":131072,"output":131072},"modalities":{"input":["text"],"output":["text"]}},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek: DeepSeek V4 Pro 0813","description":"DeepSeek V4-Pro-0813 is a 1.6T-parameter MoE model excelling at advanced reasoning, coding, and complex agentic workloads.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-08-13","last_updated":"2026-08-13","open_weights":true,"cost":{"input":1.31,"output":3.96,"cache_read":0.044},"limit":{"context":1048576,"input":1048576,"output":1048576},"modalities":{"input":["text"],"output":["text"]}},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek: DeepSeek V4 Flash 0731","description":"DeepSeek V4-Flash-0731 is an MoE model great for coding, reasoning, and agentic workloads.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-07-31","last_updated":"2026-07-31","open_weights":true,"cost":{"input":0.13,"output":0.28,"cache_read":0.07},"limit":{"context":262144,"input":262144,"output":262144},"modalities":{"input":["text"],"output":["text"]}},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek: DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a multimodal MoE model for coding, reasoning, and agentic workloads with long contexts.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-09-10","last_updated":"2026-09-10","open_weights":true,"cost":{"input":0.2,"output":0.65,"cache_read":0.03},"limit":{"context":1048576,"input":1048576,"output":1048576},"modalities":{"input":["text","image"],"output":["text"]}},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek: DeepSeek V3.1","description":"A large hybrid model that supports both thinking and non-thinking modes via prompt templates.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-08-21","last_updated":"2025-08-21","open_weights":true,"cost":{"input":0.55,"output":1.65,"cache_read":0.55},"limit":{"context":161000,"input":161000,"output":161000},"modalities":{"input":["text"],"output":["text"]}},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Meta: Llama 3.3 70B","description":"Multilingual model excelling in conversational tasks, detailed instruction-following, and coding.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2024-12-01","last_updated":"2024-12-01","open_weights":true,"cost":{"input":0.71,"output":0.71,"cache_read":0.71},"limit":{"context":128000,"input":128000,"output":128000},"modalities":{"input":["text"],"output":["text"]}}}}}