QuantumNous/new-api

What is it?

Key features

• 01Exposes OpenAI, Anthropic Messages and Gemini compatible endpoints over the same upstream models
• 02Routes requests across multiple channels with model mappings, priorities, weights and retries
• 03Manages users, groups, API keys and quotas with usage logs and cost accounting
• 04Supports OAuth/OIDC, passkeys and two-factor authentication for console login
• 05Extends asynchronous image, video and other task APIs through JavaScript plugins

Use cases

• Sharing authorized model access across a team with per-user quotas and usage tracking
• Switching upstream providers without reconfiguring every client application
• Running a private multi-model service with a web console for administrators

Quick start

1. 1

   Start a local instance with Docker (SQLite, bound to localhost)

   mkdir -p data docker run --name new-api -d --restart unless-stopped \ -p 127.0.0.1:3000:3000 \ -e TZ=Asia/Shanghai \ -v "$(pwd)/data:/data" \ calciumion/new-api:latest
2. 2

   Open http://localhost:3000 and complete the setup wizard to create the admin account

3. 3

   Add a channel with your upstream API key, configure models and pricing, then create a New API key

4. 4

   List available models with your key

   curl --fail-with-body http://localhost:3000/v1/models \ -H "Authorization: Bearer ${NEW_API_KEY}"

Things to know

• !Requires you to supply your own upstream API keys and accounts, and to comply with upstream terms.
• !Public generative AI or API resale use carries substantial legal, filing and compliance obligations.
• !Licensed under AGPL-3.0, a copyleft license with network-use source disclosure requirements.

Topics

FAQ

What is new-api?

A self-hosted AI gateway that exposes OpenAI, Claude and Gemini compatible APIs in front of many upstream model providers, with routing, keys, quotas and cost tracking. New API is a self-hosted gateway that sits between your applications and upstream model services such as OpenAI, Anthropic, Google Gemini, Azure OpenAI, AWS Bedrock, Vertex AI and DeepSeek. Clients call one consistent API, and the gateway handles conversion between OpenAI Chat/Responses, Anthropic Messages and Gemini request formats.

Is new-api free to use? What license does it have?

new-api is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.

Is new-api still maintained?

Actively maintained. The last commit to new-api was on Oct 19, 2026.

What can I use new-api for?

Sharing authorized model access across a team with per-user quotas and usage tracking Switching upstream providers without reconfiguring every client application Running a private multi-model service with a web console for administrators

How do I install new-api?

Start a local instance with Docker (SQLite, bound to localhost): mkdir -p data docker run --name new-api -d --restart unless-stopped \ -p 127.0.0.1:3000:3000 \ -e TZ=Asia/Shanghai \ -v "$(pwd)/data:/data" \ calciumion/new-api:latest Open http://localhost:3000 and complete the setup wizard to create the admin account Add a channel with your upstream API key, configure models and pricing, then create a New API key List available models with your key: curl --fail-with-body http://localhost:3000/v1/models \ -H "Authorization: Bearer ${NEW_API_KEY}"

Where can I get new-api?

The source code and installation instructions are on GitHub: https://github.com/QuantumNous/new-api