liteLLM - 2026 Pricing, Features, Reviews & Alternatives

liteLLM is an open-source AI gateway and LLM proxy that unifies 140+ providers and 1,892+ models, including OpenAI, Anthropic, Azure, Bedrock, and Vertex AI, behind a single OpenAI-compatible endpoint. Deployable via Docker or Kubernetes in self-hosted environments, it gives engineering and platform teams centralized control over model access, spend, and performance without requiring per-provider SDK integrations.

liteLLM extends that unified interface with a production-grade proxy layer featuring virtual API keys, per-key and per-team budget enforcement, rate limiting, and real-time spend dashboards. A built-in load balancer distributes requests across multiple deployments of the same model, supporting fallback routing and retry logic to maintain availability. Detailed logging integrations cover providers such as Langfuse, Helicone, Prometheus, and S3, giving observability pipelines full visibility into token consumption, latency, and error rates across every model call.

Target users include AI engineering teams, platform engineers, and enterprises standardizing LLM access across internal products. Configuration is managed through a YAML-based config file or a REST Admin API, and a web UI surfaces cost analytics and key management for non-technical stakeholders. Support tiers range from a free open-source deployment to an enterprise plan with SSO, audit logs, and dedicated support.