A web-based tool for image generation and editing using OpenAI's APIs, GPT Image Playground offers a user-friendly interface for creating and modifying images. It caters to users who prefer a graphical interface over direct API calls, providing features for text-to-image generation, reference-based editing, and parameter tuning, all while prioritizing local data storage for privacy and performance.

How It Works

Developed using React 19, TypeScript, Vite, and Tailwind CSS, the application stores all user data, including history and generated images, locally within the browser's IndexedDB. It interfaces with OpenAI's images/generations and images/edits endpoints, or the Responses API with the image_generation tool. Key features include support for multiple reference images, mask-based local editing with automatic preprocessing, and fine-grained control over parameters like resolution, quality, and format. The tool intelligently normalizes dimensions to meet API constraints and tracks actual API outputs. Its responsive design and PWA compliance ensure a seamless experience across devices.

Quick Start & Requirements

• Primary Install/Run:
  • Vercel (Recommended): One-click deployment via the provided Vercel button.
  • Docker: Pre-built images available on GitHub Container Registry.
  • Local Development: Requires Node.js and npm/yarn (npm install, npm run dev).
• Prerequisites: An OpenAI API key is mandatory. Local development requires a Node.js environment.
• Links:
  • Vercel Deployment: https://gpt-image-playground.cooksleep.dev
  • GitHub Pages Deployment: https://cooksleep.github.io/gpt_image_playground

Highlighted Details

• Local-First Data Storage: All task records and image data are stored exclusively in the browser's IndexedDB, ensuring data privacy and offline accessibility. Features automatic image deduplication and cleanup of orphaned fragments.
• Advanced Parameter Control: Offers intelligent resolution selection, automatic dimension normalization to API-compliant ranges (e.g., 16x multiples, aspect ratio limits), and tracks actual API-returned parameters, highlighting discrepancies.
• Iterative Editing Workflow: Generated images can be directly added to the reference image list for subsequent editing rounds, facilitating a fluid creative process.
• PWA Support: Enables installation as a standalone application on desktops and mobile devices, providing an immersive, native-like user experience.

Maintenance & Community

The provided README does not detail specific maintenance contributors, community channels (like Discord/Slack), or a public roadmap. Sponsorships or partnerships are not mentioned.

Licensing & Compatibility

The project is released under the MIT License. This permissive license allows for commercial use, modification, and distribution, making it compatible with closed-source applications.

Limitations & Caveats

The tool's functionality is dependent on the availability and specifications of the OpenAI API. Mask editing, while functional, is prompt-guided and does not offer pixel-level control over the edited areas. The API proxy feature, intended to bypass browser CORS restrictions, carries security risks if enabled in untrusted environments, potentially exposing the server as a proxy. The "Codex CLI Compatibility Mode" is a specialized feature for specific API interactions.