Ross ROSS = Recommend OSS · open-source software intelligence for agents

BerriAI/litellm

The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM] observed · 2026-08-28

github.com/BerriAI/litellm · homepage · Python · NOASSERTION (other) observed · 2026-08-28

Health v2 · maintenance only

91/100

  • Activity 99
  • Release rhythm 87
  • Longevity 81

Flags: no_license

How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 0
  • age_days: 1134
  • days_rel: 11
  • days_push: 7
  • n_releases_24m: 880

Full methodology

Adoption not part of the score

57340 stars · 10918 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

LiteLLM is an open-source AI gateway and Python SDK that provides a unified OpenAI-format interface to 100+ LLM providers (OpenAI, Anthropic, Bedrock, Azure, Vertex AI, vLLM, and more). It can be used as an in-code SDK or deployed as a self-hosted proxy server with virtual keys, cost tracking, budgets, load balancing, fallbacks, guardrails, and an admin UI.

Use cases

  • call openai anthropic and bedrock models through one unified api
  • self-host an llm gateway with api keys and spend tracking
  • set budgets and rate limits per team for llm usage
  • load balance and failover between multiple llm providers
  • proxy claude code or cursor through a single llm endpoint
  • add guardrails and pii masking to llm requests
  • track llm costs and token spend across an organization
  • route requests to cheaper models based on prompt complexity

When to choose

  • you use multiple LLM providers and want one OpenAI-compatible interface
  • you need centralized key management, budgets, and spend tracking for LLM access across teams
  • you want retries, fallbacks, and load balancing across model deployments
  • you need guardrails, PII masking, or observability hooks in front of LLM calls

When to avoid

  • you call a single LLM provider directly and don't need an abstraction layer
  • you need a fully permissive license - the repo has a custom license with enterprise-gated features
  • you want a lightweight client without running a proxy server or extra dependency

Facets

library · maturity active

llm-inference api-framework http-server proxy rate-limiting caching monitoring logging sdk mcp large-language-models artificial-intelligence developer-tools self-hosted apis python rust self-hosted cross-platform cli ai-gateway llm-gateway openai-compatible llm-proxy load-balancing cost-tracking guardrails virtual-keys spend-tracking fallbacks multi-provider mcp-gateway ai-agents docker

10 sources

Member repositories

RepositoryRoleHealth v2
BerriAI/litellmmain91

For agents

markdown · JSON · MCP: product_card(name="BerriAI/litellm")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem