🌌 one-api

Universal Edge LLM Gateway

One API to rule them all. Unify OpenAI, Anthropic, Google, DeepSeek, and more behind a single endpoint.

---

📑 Table of Contents

• ✨ What is this?
• 📈 Enterprise Features
• 🧠 Supported Providers
• 🏗️ System Architecture
• 🚀 Quick Start
• 📦 Docker Deployment
• ⚙️ Advanced Configuration
• 🛠️ Development

---

✨ What is this?

A dead-simple, single-binary gateway that turns dozens of LLM providers into one clean, standard OpenAI-compatible API.

Run it once and forget about provider differences. Keys, specific models, differing endpoints, and prompt formatting are all abstracted entirely away from your client code.

Why? Because juggling multiple SDKs, streaming formats, and API structures is a massive pain when trying to switch models mid-development. This just works.

---

📈 Enterprise Features

• Zero Config: Starts instantly, works out of the box, and stays up.
• Single Binary: No complex dependencies, just run it natively or via Docker.
• 100% OpenAI-Compatible: Your existing code, libraries, and SDKs (like openai-node or langchain) work completely unchanged. Just swap the baseURL.
• Lightweight & Fast: No bloated admin UIs. Features automatic response compression for reduced bandwidth and drastically lower latency.
• Streaming Support: Flawlessly streams Server-Sent Events (SSE) across all supported providers natively.
• Free & Open Source: MIT licensed forever, no hidden SaaS tiers.

---

🧠 Supported Providers

You can route requests seamlessly to any of these providers using the exact same /v1/chat/completions payload format:

• OpenAI (GPT-4o, GPT-3.5)
• Anthropic Claude (Opus, Sonnet, Haiku)
• Google Gemini (Gemini 1.5 Pro/Flash)
• Azure OpenAI
• DeepSeek
• Groq (Llama 3, Mixtral)
• Together AI
• Mistral AI
• Local inference (Ollama, LM Studio)
• +20 more API-compatible providers...

---

🏗️ System Architecture

graph LR
    Client[Your App / SDK] -->|OpenAI Format Request| Gateway[one-api Gateway]
    
    subgraph one-api Gateway
        Router[API Router] --> Translator[Payload Translator]
        Translator --> Dispatcher[Request Dispatcher]
    end
    
    Gateway -->|Native Format| P1[Anthropic API]
    Gateway -->|Native Format| P2[Google Gemini API]
    Gateway -->|Native Format| P3[OpenAI API]
    Gateway -->|Native Format| P4[Local Ollama]

Loading

---

🚀 Quick Start

1. Clone & Install

git clone https://github.com/shenald-dev/one-api
cd one-api
npm install

2. Configure Environment

cp .env.example .env
# Open .env and add your provider API keys

3. Run the Server

npm start

4. Test the Endpoint

Point any curl command or SDK to http://localhost:3000/v1:

curl -H "Content-Type: application/json" \
     -d '{"model":"claude-3-haiku-20240307","messages":[{"role":"user","content":"Hello!"}]}' \
     http://localhost:3000/v1/chat/completions

Notice how you passed an Anthropic model to an OpenAI-compatible endpoint!

---

📦 Docker Deployment (Recommended)

The most robust way to run one-api in production:

docker run -d \
  -p 3000:3000 \
  -v $(pwd)/data:/app/data \
  -e OPENAI_API_KEY=sk-... \
  -e ANTHROPIC_API_KEY=sk-... \
  shenald/one-api:latest

---

⚙️ Advanced Configuration

All routing is controlled via environment variables in the .env file:

# Server Configuration
PORT=3000
ENABLE_COMPRESSION=true

# Provider Keys
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-...
GOOGLE_API_KEY=...
DEEPSEEK_API_KEY=...
GROQ_API_KEY=...
OLLAMA_BASE_URL=http://localhost:11434

No databases required. No migrations. State is entirely managed via configuration.

---

🛠️ Development

Want to add a new provider adapter?

# Install deps
npm install

# Run dev with hot reload
npm run dev

# Run the test suite
npm test

# Build Docker image locally
docker build -t shenald/one-api .

---

🙋‍♂️ About

Built by a vibe coder who got tired of rewriting API integrations.
If it's useful, star it ⭐ — if not, open an issue and tell me why.

Keep it simple. 🧘