Model Compression

This tool applies ALL optimizations in a single pass: mesh simplification + texture resize/compression + Draco encoding. It is the most comprehensive option for maximum file size reduction.

Mesh simplification (powered by the Meshopt library) actually reduces the number of faces/vertices in your model — higher strength means fewer polygons, smaller files, but visible quality loss. Draco encoding compresses the existing geometry data without changing face count, achieving 50-90% size reduction with nearly no visual difference. For best results, use both together: apply Meshopt simplification first, then Draco encoding.

Results vary widely: simple models may see 30-50% reduction, while complex textured models can achieve 70-90% reduction when all options are enabled.

For web deployment, enabling all options (simplify + texture compress + Draco) gives the smallest files. For archival purposes, you may want to skip simplification to preserve full detail.

WebP offers the best compression ratio (~30% smaller than JPEG at similar quality) and is supported by all modern browsers. Use JPEG for maximum compatibility if needed.

Draco compression only works with GLB/GLTF format, so the output is always GLB. To convert it back to OBJ, STL, or other formats, use our GLB format converter tool.

KTX2 is a GPU-optimized texture format that enables hardware-accelerated decoding. Note that KTX2 textures cannot be opened by most image viewers or standard 3D software. You'll need compatible tools such as Three.js (with KTX2Loader), Babylon.js, or dedicated GPU texture tools to view and use them.