LiteLLM

Call 100+ LLMs using the OpenAI Input/Output Format.

• Translates inputs to provider endpoints (/chat/completions, /embeddings, /images, etc.)

• Consistent output format regardless of provider

• Retry/fallback logic across multiple deployments (Azure/OpenAI) via Router

• Track spend and set budgets per project via LiteLLM Proxy Server

• Docs: https://docs.litellm.ai/

• GitHub: https://github.com/BerriAI/litellm

Related

• vLLM
• Ollama
• OpenRouter
• Claude_Developer_Platform
• Google_AI_Studio
• Cohere