Freemiumcoding

LiteLLM

Put models, agents, and MCP servers behind one governed AI gateway.

Open-source, self-hosted AI gateway and proxy for routing, governing, and observing access to many LLM providers through compatible APIs.

4.0

Try LiteLLM Free →

Opens official website

AI gatewayLLM proxymodel routingcost controlself-hosted

Live change feed

Recent verified changes

Shared with the homepage and tool pages

newlitellmJul 28, 2026

LiteLLM 1.94.0 adds signed Docker images and stronger proxy controls

LiteLLM now signs its Docker images with Cosign so teams can verify container provenance before deployment; the release also tightens proxy authorization and sanitizes guardrail scan results before logging.

Verify official source →

About LiteLLM

LiteLLM is an open-source, self-hosted AI gateway rather than a hosted model subscription. Its proxy gives applications a unified interface for more than 100 model providers and supports virtual keys, budgets, rate limits, spend tracking, routing, fallbacks, logging, and metrics. The open-source gateway is free to self-host in production. Enterprise adds commercial governance, security, deployment, and support capabilities including SSO and SCIM, OIDC or JWT authentication, audit logs, secret-manager integration and key rotation, organization administration, multi-region control-plane capabilities, air-gapped deployment options, and support with SLAs. LiteLLM Enterprise is quoted annually according to gateway request capacity, deployment architecture, and support needs; LiteLLM states that this license is not priced per token. The model-provider API charges and the customer's own infrastructure costs remain separate from LiteLLM's $0 open-source license.

✅ Pros

+Open-source gateway is free to self-host, including for production use
+Unified OpenAI-compatible interface for more than 100 model providers
+Supports virtual keys, budgets, rate limits, spend tracking, routing, fallbacks, logging, and Prometheus metrics
+Works with common application clients and SDKs through compatible endpoints
+Enterprise adds SSO, SCIM, OIDC or JWT, audit logs, secret-manager integrations, key rotation, and organization administration
+Enterprise supports self-hosted, air-gapped, and multi-region deployment paths with commercial support options

❌ Cons

−Open-source use is self-hosted, so operations, upgrades, security, and infrastructure remain the customer's responsibility
−LiteLLM does not include model inference: model-provider API charges remain separate
−Cloud, database, cache, observability, and networking costs are separate from the $0 software license
−Enterprise license pricing is not publicly listed and requires a sales quote
−A production deployment can require supporting components such as a database, Redis, secret management, ingress, and monitoring depending on the chosen architecture
−Enterprise-only features and support require a commercial agreement

✅ Best For

Engineering teams unifying several LLM providers behind one APITeams that need keys, budgets, rate limits, routing, fallbacks, and model-spend visibilityOrganizations deploying an internal or self-hosted AI gatewayPlatform teams standardizing access across OpenAI, Anthropic, Bedrock, Vertex AI, Azure, local models, and compatible endpointsRegulated or large organizations needing enterprise identity, audit, air-gapped, or support requirements

Quick Info

Pricing: freemium
Category: coding
Rating: 4.0 / 5.0
Last Reviewed: 2026-08-24T00:00:00+00:00

💰 Pricing Breakdown

Official pricing as of 2026-08-24: LiteLLM Open Source is $0 and free forever when self-hosted, including production use. Enterprise is annual, custom-quoted by gateway request capacity, deployment architecture, and support needs; LiteLLM says it is never priced per token. An instant 30-day Enterprise trial key is offered without a credit card. Model-provider API spend and self-hosted infrastructure costs are separate and are not LiteLLM subscription charges.

View current pricing →

Try LiteLLM Free →

More coding tools