LiteLLM
AI InfrastructureAn open-source LLM gateway that gives you one consistent API and proxy across 100+ model providers, with key management and spend tracking.
Overview
LiteLLM solves a mundane but real problem: every model provider has a slightly different API, and hard-coding one of them is how you end up unable to switch. It exposes a single OpenAI-compatible interface across providers, and its proxy server adds centralized API-key management, per-team budgets, rate limiting, and spend logging. For enterprises that is the point — one control plane for all model traffic, with cost attribution and the freedom to route or fail over between providers. Lightweight to adopt, and a sensible default early.
Pros & Cons
Pros
- One consistent API across 100+ providers
- Proxy adds key management, budgets, and spend logging
- Avoids provider lock-in
- Lightweight to adopt
Cons
- Self-hosting the proxy is your operational burden
- A gateway is one more hop to monitor
- Advanced governance features are in the paid edition
Workflows that use LiteLLM
Get a new AI workflow each week — many feature LiteLLM and other tools in this category.