Generates custom Gridfinity-compatible 3D printable bins using AI to trace tool outlines from photographs. It targets makers and 3D printing enthusiasts, offering an efficient method to design bespoke storage solutions precisely tailored to individual tools, significantly reducing manual design effort.

How It Works

Users upload a top-down photo of tools on paper. Tracefinity employs AI models—either local (IS-Net, BiRefNet Lite) or Google Gemini API—to automatically trace tool outlines, converting them into editable polygons. These traced tools can be managed in a library, grouped into projects for workspace planning, and used to design custom bins. The system supports snap-to-grid layouts, finger holes, and text labels before exporting designs as STL/3MF for 3D printing.

Quick Start & Requirements

A live demo is available at tracefinity.net. For self-hosting, Docker is recommended:

docker run -p 3000:3000 -v ./data:/app/storage ghcr.io/tracefinity/tracefinity

This uses local tracing by default; set GOOGLE_API_KEY for Gemini. Building from source requires Python 3.11+, Node.js 20+, and pnpm (make dev). Local tracing models have RAM needs from 2GB (IS-Net) to 8GB+ (BiRefNet Lite). Optional GPU acceleration via ONNX Runtime is supported.

Highlighted Details

• AI-powered tool tracing (local models or Gemini API).
• Tool library and project management for storage planning.
• Gridfinity-standard bin generation with magnet holes, stacking lips, finger holes, and text labels.
• Tool editor for refining outlines, smoothing curves, and precise placement.
• Exports include STL, 3MF (multi-color), and SVG.
• Automatic bed splitting for large bins and configurable cutout clearances/chamfers.

Maintenance & Community

The provided README does not detail specific maintenance contributors, sponsorships, or community channels.

Licensing & Compatibility

Released under the MIT License, permitting commercial use and integration within closed-source projects.

Limitations & Caveats

Some Gemini models are "preview" versions. Local tracing performance and RAM requirements vary significantly by model (up to 8GB+ for BiRefNet Lite). GPU acceleration requires additional setup.