This project provides a local, desktop-like web application for generating images using OpenAI's GPT models, mimicking the ChatGPT/Codex workflow. It targets users seeking an iterative image creation studio with advanced features like history, branching, and cleanup, offering a free OAuth-based authentication path alongside paid API key support.

How It Works

ima2-gen operates as a Node.js application, accessible via a local web UI or CLI. Its core innovation lies in offering dual authentication: a default, free path using Codex OAuth, and an alternative using a configured OpenAI API key. The system supports multiple generation workflows: Classic for single results, Node for exploring divergent image paths, Multimode for parallel generation, and Canvas Mode for detailed image editing and background cleanup. It features a local, session-aware gallery and allows prompt library imports from various sources, enhancing local iteration and asset management.

Quick Start & Requirements

• Primary install/run: Execute npx ima2-gen serve to start the local server. For initial OAuth setup, run npx @openai/codex login.
• Prerequisites: Node.js environment.
• Links:
  • Live site: lidg-jun.github.io/ima2-gen
  • CLI Reference: docs/CLI.md
  • API Reference: docs/API.md

Highlighted Details

• Dual authentication: free Codex OAuth or paid OpenAI API key.
• Advanced workflows: Node mode for branching, Multimode for parallel generation, Canvas Mode for detailed cleanup.
• Local gallery with session-aware history and prompt library imports.
• Support for multiple GPT models (gpt-5.4-mini, gpt-5.4, gpt-5.5) with configurable quality and size.

Maintenance & Community

No specific details regarding community channels (e.g., Discord, Slack), active contributors, or sponsorships were found in the provided README.

Licensing & Compatibility

• License: MIT.
• Compatibility: The MIT license permits commercial use and integration with closed-source projects.

Limitations & Caveats

The "Card News Mode" is experimental and not intended for stable use. HEIC/HEIF image formats are not supported for reference images via the browser path. Certain advanced models like gpt-5.5 may require updated tooling or backend support. Windows users may encounter file lock issues during global package updates.