One API
Self-hosted LLM API gateway with an OpenAI-compatible API, Ollama support and access controls. Runs as a single executable or in Docker under MIT.
One API is a self-hosted LLM API gateway for developers and teams that want to share model access across apps or users. It puts cloud providers and Ollama behind an OpenAI-compatible API, so clients can use one endpoint across different backends. It's open source under MIT and runs on your own server as a single executable or in Docker.
Supported services include OpenAI, Azure OpenAI, Anthropic Claude, Google Gemini and DeepSeek. Ollama provides a connection to models you run locally. The gateway and its stored data live on your server; requests routed to cloud providers leave that server for processing and require internet access.
Access management is a central feature. You can issue separate tokens with expiration dates, model permissions and IP restrictions, then organize users and provider connections into groups. Usage records help administrators track consumption without handing each user the upstream provider's key.
The gateway balances requests across multiple provider connections and retries failed requests automatically. It supports streamed responses and image-generation APIs, and model mapping lets administrators redirect a requested model name to another model. Multi-server deployment is also supported.
An English web interface handles administration, while a management API lets other software manage the service. The user portal supports custom branding and pages, with email, GitHub and Feishu login options.
