Why convert OBJ to GLB?
OBJ is universally supported but dated: it is uncompressed ASCII, stores vertices without indexing, and splits a model across an .obj, an .mtl, and a pile of image files. That is slow to download and slow to upload to the GPU.
GLB packs geometry, materials, and textures into one compact binary that loads in a single request — ideal for web, AR, and game runtimes.
What Zlim optimizes
Zlim welds duplicated vertices into an indexed mesh, quantizes attributes, and applies Draco compression to shrink the geometry payload. Textures referenced by the MTL are transcoded to KTX2 and mipmapped, cutting GPU texture memory — the half of the budget OBJ-to-GLB exporters usually ignore. Draw calls drop as compatible meshes and materials are merged.
zlim optimize model.obj -o model.glb -p balanced # bundles model.obj + .mtl + textures → model.glbFAQ
Do I need to upload the MTL and textures too?
Yes — include the .mtl and referenced image files (a zip or multi-file upload). Zlim resolves the material references and bundles everything into the single GLB.
OBJ has no animation — what's the win?
Indexing, quantization, Draco geometry compression, and KTX2 textures. For typical OBJ exports that is a 70–80% size reduction plus far fewer draw calls.
Convert OBJ to GLB now
Start free — 25 optimizations a month. No card required.