This project provides a reverse-engineered API wrapper for Jimeng (China) and Dreamina (international) AI text-to-image, image-to-image, and video generation services. It targets developers and researchers seeking programmatic access to these AI features, offering a stable and enhanced interface with advanced error handling and deployment options.

How It Works

The API is built upon reverse-engineering the official Jimeng and Dreamina platforms, enabling access to their AI generation capabilities. It implements core features like text-to-image, image-to-image synthesis, and AI video generation. Key architectural choices include a "Smart Polling" mechanism to optimize generation efficiency and a "Unified Exception Handling" system with automatic retries, aiming for increased stability and robustness over direct interactions with the official services.

Quick Start & Requirements

• Primary Install: Recommended installation is via Docker: docker run -d --name jimeng-api -p 5100:5100 ghcr.io/iptag/jimeng-api:latest. Alternatively, clone the repository, install dependencies (npm install), and run (npm run dev). Docker Compose is also supported.
• Prerequisites: Node.js 18+ and npm/yarn are required for direct installation. Docker is optional but recommended. A valid YOUR_SESSION_ID obtained from the Jimeng/Dreamina website is mandatory for API authentication.
• Links: No external quick-start guides or demos are provided beyond the README.

Highlighted Details

• Multi-modal AI: Supports text-to-image, image-to-image, and text-to-video/image-to-video generation across multiple models.
• Resolution & Ratio Control: Offers granular control over image output dimensions, supporting resolutions from 1k to 4k and various aspect ratios (e.g., 1:1, 16:9, 9:16).
• Claude Code Skill: Integrates directly with Claude Code, allowing image generation within conversations with automatic saving and format conversion (WebP to PNG).
• Docker Optimized: Features a multi-stage build for a small image size (170MB), runs as a non-root user for enhanced security, and includes health checks.

Maintenance & Community

No specific details regarding maintainers, active community channels (like Discord/Slack), or sponsorship are provided in the README. Users are encouraged to file issues for problems.

Licensing & Compatibility

• License: GPL v3.
• Compatibility: As a GPL v3 licensed project, it is copyleft. Use in commercial or closed-source applications requires careful consideration of the license terms and potential obligations.

Limitations & Caveats

This project is strictly for "research and educational purposes only" and explicitly warns against pressuring official servers, with potential consequences including account bans or legal issues. It does not accept financial donations. The functionality relies on valid session IDs from the official services, which are subject to change or revocation by the providers.