Summary

This project offers a self-hostable API gateway emulating OpenAI's chat completion and Anthropic's Messages API formats. It enables users to act as AI assistants, integrate with diverse AI clients, or serve as a mock LLM for agent development, featuring a web console for configuration and request assembly.

How It Works

ChatAPI employs a Flask backend and a React/Vite/Ant Design frontend, leveraging SQLite for data persistence. It exposes OpenAI-compatible endpoints (/v1/chat/completions, /v1/responses) and an Anthropic-compatible /messages endpoint. Key functionalities include user authentication, session management, message history storage, and advanced features like assembling Tool Calling requests via its web UI and configuring automated, conditional, or time-based responses.

Quick Start & Requirements

• Installation: Clone repo, build frontend (cd ./frontend && npm i && npm run build), configure .env (min: CHATAPI_USERNAME, CHATAPI_PASSWORD, CHATAPI_SESSION_SECRET), start backend (cd ./backend && uv sync && uv run main.py dev).
• Prerequisites: Node.js (frontend build), Python (uv for backend), SQLite.
• Configuration: Environment variables in .env manage credentials, secrets, database path, host, port, and optional TLS/email settings.
• Docs: README serves as primary documentation.

Highlighted Details

• Emulates OpenAI /v1/chat/completions, /v1/responses, and Anthropic /messages APIs.
• Web console facilitates Tool Calling request assembly and automated/conditional reply rule configuration.
• Supports user authentication (optional TOTP) and message persistence for debugging.
• Optional ntfy integration for message push notifications.

Maintenance & Community

No specific details on maintainers, community channels, or roadmap were found in the provided text.

Licensing & Compatibility

The license type is not explicitly stated in the provided README content.

Limitations & Caveats

The ntfy message push feature presents a potential Server-Side Request Forgery (SSRF) risk if configured to access internal network addresses (127.0.0.1, localhost, internal IPs) without adequate trust; mitigation options are available via configuration settings (e.g., "Admin only" or "Closed"). Secure operation necessitates careful environment variable setup.