Mesh optimization — decimate and compress glTF / STL in the browser

1. Upload

GLB, glTF (embedded), or STL. Drag & drop or browse. Nothing is uploaded to a server.

Drop a model here or click to choose

.glb / .gltf / .stl · For OBJ, use the CLI locally

What this tool does

It takes a 3D model, runs a mesh simplification pass (fewer triangles within a surface error you control), then packs the result as a compressed glTF binary (.glb) suitable for websites and WebGL viewers. STL uploads are converted through the same pipeline. Nothing is uploaded to a server — decoding and optimization use WebAssembly on your machine.

Fewer triangles

Uses meshoptimizer simplification after weld/dedup so meshes stay cleaner while you dial in how much geometry to keep (ratio or target count) and how much shape deviation is acceptable.
Smaller GPU payloads

Writes EXT_meshopt_compression on the geometry buffers so downloads and GPU uploads are smaller than raw floats, with presets for medium (faster decode) or high (tighter packing).
Preview & compare

Loads your file with the same Three.js path as many production sites (including Draco and meshopt decoders when present). After optimizing, you get side‑by‑side previews plus triangle, vertex, and file-size stats.
Exports

Download an optimized .glb (full glTF scene) or .stl that matches the framed optimized preview (centered and scaled for the viewer — useful for quick STL handoff).

How the pipeline works (browser)

Input is read as glTF or converted from STL into a minimal glTF document (positions, indices, default material).
glTF Transform runs dedup, weld, MeshoptSimplifier simplify, then meshopt buffer compression, and prunes unused data.
Preset web-fast targets a moderate reduction with medium meshopt; web-small pushes stronger simplification and high meshopt. (Smallest possible transfers with KHR_draco_mesh_compression require the Node CLI — Draco encoding is not available in this web build.)

Also in the repo

The mesh-optimize command-line tool adds OBJ simplification (Fast Quadric) and optional Draco for glTF. See the project on GitHub for install and presets.

mesh-optimization

1. Upload

2. Inspect & preview

Original

3. Preset & compression

web-fast

web-small

Decimation

Quality & compression

4. Result

What this tool does

Fewer triangles

Smaller GPU payloads

Preview & compare

Exports

How the pipeline works (browser)

Also in the repo