This project provides a reverse-engineered, OpenAI-compatible API for Jimeng AI's image generation capabilities, targeting developers and researchers who want to integrate advanced image generation into their applications without direct reliance on Jimeng's official SDKs. It offers zero-configuration deployment and supports multiple account tokens for increased usage.

How It Works

The service acts as a proxy, translating requests formatted for OpenAI's API into the proprietary calls required by Jimeng AI. It extracts a sessionid from Jimeng's web interface to authenticate requests, effectively mimicking a logged-in user. This approach allows for broad compatibility with existing OpenAI-compatible clients and frameworks.

Quick Start & Requirements

• Docker: docker run -it -d --init --name jimeng-free-api -p 8000:8000 -e TZ=Asia/Shanghai vinlic/jimeng-free-api:latest
• Prerequisites: A server with a public IP and port 8000 open. Requires a Jimeng AI sessionid obtained from browser cookies.
• Setup: Minimal setup time if Docker is available.
• Docs: Jimeng AI Official

Highlighted Details

• OpenAI API compatibility for chat completions and image generation.
• Supports multiple sessionid tokens, concatenated with commas, for load balancing across accounts.
• Offers various Jimeng models including jimeng-3.0, jimeng-2.1, jimeng-2.0-pro, etc.
• Includes deployment guides for Docker, Docker-compose, Render, and Vercel.

Maintenance & Community

The project is part of a larger "free-api" series by LLM-Red-Team, which includes wrappers for other AI models like Kimi.ai, StepChat, and Qwen. Community interaction channels are not explicitly mentioned in the README.

Licensing & Compatibility

The project itself appears to be open-source, but the underlying Jimeng AI service is accessed via reverse-engineering. The README explicitly states: "Reverse API is unstable, recommend experiencing features through the official Jimeng AI... Avoid the risk of being blocked!" and "For self-use only, prohibited from providing services to others or commercial use."

Limitations & Caveats

This is a reverse-engineered API, making it inherently unstable and subject to breakage if Jimeng AI changes its internal structure or implements stricter anti-scraping measures. The README strongly warns against commercial use and emphasizes that users bear all risks associated with potential account blocking. Vercel deployments may encounter 504 timeouts due to the 10-second response limit.