DAE to GLB converter: the skeleton and the skin weights kept
Drop a .dae and get one .glb file, converted in Blender on our server. The trip keeps the skeleton and the skin weights, the shape keys and the materials and the textures; nothing the source holds is lost on the way. Free, no account, nothing to install.
How do I convert DAE to GLB?
Drop the file on this page, wait while the server converts it, and download one .glb file.
- 1Choose a .dae, or a zip of the folder that holds it with its textures. Files up to 50 MB.
- 2The converter opens it in Blender on our server and writes GLB. Most files take under a minute; the page shows the result in 3D when it is done.
- 3Download one .glb file. The upload and the result are deleted 3 days later.
- In:
- .dae (or a zip with its textures)
- Out:
- one .glb file
- Keeps:
- the skeleton and the skin weights, the shape keys and the materials and the textures
- Cost:
- free, no account
- The skeleton is written, the motion curves are not.
What survives the trip
A Collada file states its unit and its up axis and both are honoured. It names its images by path and does not contain them, so zip the .dae with its texture folder. A file that names an image by an absolute path, a web address, a path leading out of the folder, or that pulls in another Collada document, is refused: re-export it as one document with relative paths.
The result is one .glb with the geometry, the materials, the skeleton and the images packed inside it, in metres and Y-up as glTF defines. Nothing has to travel beside it.
DAE and GLB do not hold the same things, and the page says so before you upload: the trip keeps the skeleton and the skin weights, the shape keys and the materials and the textures, and nothing the source holds is lost on the way. GLB is what web viewers, three.js, model-viewer, Godot, Unity and Unreal importers read without a plug-in.
What this conversion cannot do
Measured on real files, not assumed. Everything below is a property of the formats or of the pipeline, and knowing it beforehand is cheaper than discovering it in your 3D tool.
Collada importers disagree with each other
The reader handles files written by Blender and by most current exporters. Files from older Maya and 3ds Max plug-ins, SketchUp or game rippers vary a great deal. Open the result in a viewer before relying on it.
Vertex counts go up, triangle counts do not
glTF stores one vertex per unique position, normal and texture coordinate, so a hard edge or a UV seam splits a shared vertex into two or three. A viewer that counts vertices will show more than your 3D tool did; the triangle count is the one that means the same on both sides.
Animation is not carried
The skeleton is written, the motion curves are not: reading a long clip costs more memory than the converter has, so the importer is asked to skip it. A file that arrives with a walk cycle leaves with the same rig in its bind pose. A motion you want on an avatar is the FBX to VRMA converter's job.
Uploads stop at 50 MB
A larger file is refused before anything is uploaded. Decimate a dense mesh or shrink the textures in your 3D tool first; neither changes the format.
Frequently asked questions
Will the model keep its real size?
Yes. Unlike STL or PLY, a Collada file states its unit and its up axis, and both are applied: a model declared in centimetres and Y-up arrives at the right size and upright.
Do I get a .glb or a .gltf?
A .glb, the single-file form. Every viewer and engine that reads .gltf reads .glb, and there is no folder to keep together.
Is anything uploaded, and for how long?
Yes: the file you drop goes to our server in New York, is converted there, and both the upload and the result are deleted 3 days after the conversion. Nothing is shared with anyone.
Can I look at the DAE first?
Yes, without uploading anything: the DAE viewer draws it in your browser. Open the DAE viewer →
I need a VTubing avatar, not a plain model
Then GLB is a stop on the way: VSeeFace, Warudo and VNyan load .vrm. The FBX to VRM converter takes the same rigged file, maps its skeleton onto the VRM humanoid and carries the materials over. FBX to VRM converter →
The other format pairs
All 50 are free, all run the same pipeline, and all say what they drop.
GLB to FBX
GLB in, FBX out, free: one file.
GLB to OBJ
GLB in, OBJ out, free: zipped with its files.
FBX to GLB
FBX in, GLB out, free: one file.
OBJ to GLB
OBJ in, GLB out, free: one file.
STL to GLB
STL in, GLB out, free: one file.
PLY to GLB
PLY in, GLB out, free: one file.
DAE to FBX
DAE in, FBX out, free: one file.
DAE to OBJ
DAE in, OBJ out, free: zipped with its files.
USDZ to GLB
USDZ in, GLB out, free: one file.
GLB to STL
GLB in, STL out, free: one file.
GLB to PLY
GLB in, PLY out, free: one file.
GLB to DAE
GLB in, DAE out, free: zipped with its files.
GLB to USDZ
GLB in, USDZ out, free: one file.
DAE to STL
DAE in, STL out, free: one file.
DAE to PLY
DAE in, PLY out, free: one file.
DAE to USDZ
DAE in, USDZ out, free: one file.
3MF to GLB
3MF in, GLB out, free: one file.
GLB to 3MF
GLB in, 3MF out, free: one file.
FBX to DAE
FBX in, DAE out, free: zipped with its files.
OBJ to DAE
OBJ in, DAE out, free: zipped with its files.
STL to DAE
STL in, DAE out, free: zipped with its files.
PLY to DAE
PLY in, DAE out, free: zipped with its files.
USDZ to DAE
USDZ in, DAE out, free: zipped with its files.
Only want to look at the file? GLB viewer, FBX viewer and OBJ viewer all run on your own device — nothing is uploaded.
Is your model going into VTubing?
Converting because the model is going into VTubing? GLB is not the format those apps read. The same rigged file becomes a VRM with its humanoid bones mapped and its materials converted, in one step on the FBX to VRM converter.
Open the VRM converter