Repolode
Submit
theopenco

theopenco/llmgateway

An open-source API gateway that routes LLM requests across multiple providers through an OpenAI-compatible interface, with key management and usage analytics.

Open-source alternative to:OpenRouter

View on GitHubWebsite
Stars
1,729
Forks
201
Open issues
87
Last commit
yesterday

What is it?

LLM Gateway sits between your applications and large language model providers such as OpenAI, Anthropic and Google Vertex AI. Applications send requests to one endpoint, and the gateway forwards them to the chosen provider, so you do not need a separate integration for each service.

API keys for different providers are managed in one place. The gateway tracks requests, token usage, response times and costs, and includes performance comparisons across models to help evaluate cost-effectiveness.

It can be used as a hosted service at llmgateway.io or self-hosted with Docker, using a unified container that bundles PostgreSQL and Redis. The codebase is a TypeScript monorepo with a Hono backend, Next.js dashboards and a Drizzle ORM schema. The core is AGPLv3, and certain enterprise features in the ee/ directory are under a separate commercial license.

Key features

  • 01Exposes an OpenAI-compatible API for routing to multiple providers
  • 02Centralized management of API keys for different providers
  • 03Tracks token usage, costs, and response times per request
  • 04Compares performance and cost across models
  • 05Can be self-hosted with Docker or used as a hosted service

Use cases

  • Switching an existing OpenAI-based application to other providers with minimal code changes
  • Tracking LLM spending and token usage across a team or multiple projects
  • Comparing model cost and latency before choosing a default model
  • Running a self-hosted LLM gateway that keeps request data under your own control

Quick start

  1. 1

    Install dependencies and set up the development environment (requires Docker)

    pnpm i && pnpm run setup
  2. 2

    Start the development servers

    pnpm dev

Things to know

  • Core code is licensed under AGPLv3, a copyleft license with obligations for networked services.
  • Commercial features in the ee/ directory require a paid Enterprise license.
  • Development setup depends on Docker services for the database and related components.

In collections

Claude Code tools, skills & pluginsOpenTelemetry & observability platformsMicroservice frameworks & service infrastructure

Topics

#llm#ai-gateway#openai-compatible#api-gateway#observability#rate-limiting#typescript#ai#analytics#anthropic

FAQ

What is llmgateway?
An open-source API gateway that routes LLM requests across multiple providers through an OpenAI-compatible interface, with key management and usage analytics. LLM Gateway sits between your applications and large language model providers such as OpenAI, Anthropic and Google Vertex AI. Applications send requests to one endpoint, and the gateway forwards them to the chosen provider, so you do not need a separate integration for each service.
Is llmgateway free to use? What license does it have?
llmgateway is open source under the Other license. Check the license terms before using it in commercial products.
Is llmgateway still maintained?
Actively maintained. The last commit to llmgateway was on Nov 6, 2026.
What can I use llmgateway for?
Switching an existing OpenAI-based application to other providers with minimal code changes Tracking LLM spending and token usage across a team or multiple projects Comparing model cost and latency before choosing a default model Running a self-hosted LLM gateway that keeps request data under your own control
How do I install llmgateway?
Install dependencies and set up the development environment (requires Docker): pnpm i && pnpm run setup Start the development servers: pnpm dev
Where can I get llmgateway?
The source code and installation instructions are on GitHub: https://github.com/theopenco/llmgateway

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects