FBX Compressor

Compress an FBX for real-time use: the model is converted to an optimized GLB — quantized geometry, deduplicated data, WebP textures — usually a fraction of the FBX's size.

FBX in → optimized GLB out

Drag & drop your FBX file here

Compress FBX models for real-time use — output is an optimized GLB, the compressed runtime format engines and viewers load. Drop texture images alongside the FBX to include them.

Or click to browse

FBX (+ PNG, JPG, TGA textures)

How it works

1

Upload FBX

Textures alongside if not embedded

2

Tune compression

Mesh quantization, texture quality & size

3

Download GLB

Smaller file, verified loadable

How It Works

1

Upload an FBX

Drop in the model, with its texture images alongside if they aren't embedded.

2

Tune the compression

Choose texture quality and size limits, quantization, or meshopt for the smallest output.

3

Download optimized GLB

A verified, engine-ready file — typically a fraction of the FBX's size.

Frequently Asked Questions

There is no in-browser FBX writer, and more importantly, GLB is what you actually ship: engines, Three.js, Babylon, and viewers load it natively, and it supports real compression (quantization, meshopt, WebP textures) that FBX doesn't. Keep the FBX as your editing master; ship the GLB.

It depends on what dominates the file. Texture-heavy models often shrink 60–90% from WebP re-encoding and downscaling alone; geometry-heavy models benefit most from quantization and meshopt. The result panel shows the exact before and after.

Yes — skeletons and animation clips convert and compress along with the meshes, and the output is re-imported after compression to verify it still loads.

No. Conversion and compression run entirely in your browser via Three.js and glTF-Transform — nothing leaves your machine.