# A2Agent

> A2Agent is an API gateway that exposes Chinese large language models — GLM (Zhipu), Kimi (Moonshot), DeepSeek, Qwen (Alibaba) and MiniMax — through OpenAI-, Anthropic- and Gemini-compatible endpoints. Any client that already speaks one of those three protocols works unchanged: point its base URL here and use an A2Agent key.

## API access

- Base URL: https://api.a2agent.me
- Authentication: `Authorization: Bearer <your API key>` on every request
- OpenAI-compatible: `POST /v1/chat/completions`, `POST /v1/responses`, `GET /v1/models`
- Anthropic-compatible: `POST /v1/messages`, `POST /v1/messages/count_tokens`
- Gemini-compatible: `POST /v1beta/models/{model}:generateContent`
- Works with Claude Code (Anthropic surface), Codex CLI and Gemini CLI without a proxy shim.

## Full context

- [Full listing with live model pricing](https://a2agent.me/llms-full.txt): quickstart, the complete model price table in USD per 1M tokens, and every published article.

## Pages

- [Home](https://a2agent.me/): what the gateway does and who it is for.
- [Pricing](https://a2agent.me/pricing): per-token billing and model prices.
- [Models](https://a2agent.me/models): every model available through the gateway, grouped by platform.
- [Docs](https://docs.a2agent.me): connecting Claude Code, Codex CLI and Gemini CLI.
- [Status](https://a2agent.me/status): live upstream health per channel.
- [Blog](https://a2agent.me/blog): articles. Append `.md` to any article URL for its markdown source.
- [About](https://a2agent.me/about): who runs the service.
- [What is A2Agent](https://a2agent.me/what-is-a2agent): the gateway, who it is for, and how it relates to A2A.

## Optional

- [Terms of Service](https://a2agent.me/terms)
- [Privacy Policy](https://a2agent.me/privacy)
- [Refund Policy](https://a2agent.me/refund-policy)