One API is an open-source LLM API management and distribution system with 37,000+ stars on GitHub that lets you access many large-model providers through the standard OpenAI API format.
Every model provider ships a different API, key system and billing model. One API sits in front of them all: deploy the single binary (or Docker image), add your upstream keys as channels, and your applications talk to one OpenAI-compatible endpoint while One API translates, routes and keeps the books. It supports an unusually long provider list — OpenAI and Azure OpenAI, Anthropic Claude, Google Gemini, DeepSeek, ByteDance Doubao, Baidu ERNIE, Alibaba Qwen, iFlytek Spark, Zhipu ChatGLM, Moonshot, MiniMax, Groq, Ollama and more. Around the gateway it provides the management layer teams actually need: user and token management, per-key quotas, channel grouping and failover, usage statistics and redemption codes for distributing credit. An English UI is available alongside the original Chinese. It remains actively maintained, with updates into 2026.
One API is free and open source under the MIT license for self-hosting. You only pay the upstream model providers whose keys you plug into it.
Developers, teams and communities consolidating many LLM accounts into one endpoint — labs sharing keys, internal AI platforms and anyone reselling or rationing access under proper authorization.