Free online DAE viewer
Drop a Collada .dae with its texture images, the folder, or a .zip. If the file has an animation it plays; if it is rigged without one, it stands in its bind pose. Nothing is uploaded.
How to open a DAE file
- Drop the .dae together with its textures. A Collada file only names its images, so select the model and the image files together, drop the whole folder, or drop the .zip the model came in.
- The model appears the right way up and at its real scale: the file states its up axis and how long its unit is in metres, and both are applied. The panel shows what it said.
- Check the texture list for images the file asks for that were not among your files, and use the play / pause button if the file brought an animation.
What Collada is for
Collada is an XML interchange format from the mid-2000s, kept by the Khronos Group, meant to move a whole scene between 3D programs: meshes, materials, textures by file path, skeletons and skin weights, animation, cameras and lights. It is text, so a .dae can be opened in an editor and read.
It is an old format now. SketchUp still exports it, Second Life and OpenSimulator take mesh uploads in it, and a great deal of game-modding tooling from the 2010s speaks nothing else. Newer pipelines have moved to glTF, and Blender removed its Collada importer and exporter in version 5.0 — which is one reason to want a viewer that does not need installing.
For avatars: a .dae can hold a rigged humanoid, so unlike STL or PLY it is a possible starting point. VRM tools do not read Collada, though — the route is DAE to GLB first, then GLB to VRM.
What this viewer does not do
It plays the first animation clip only; there is no clip list and no timeline. Lights and cameras stored in the file are left out and the viewer’s own light is used, so a scene will not look as it did in the program that wrote it. Kinematics and physics sections are not shown. Materials are drawn with the classic Phong or Lambert model the format describes — colour, texture, transparency — not with a physically based shader. Textures in formats a browser cannot decode (.dds, .tif) are listed as present but not drawn; .tga is decoded.
Frequently asked questions
The model is white or grey. Where are the textures?
A .dae does not contain images; it names them by path, often an absolute path on the author's disk. Drop the image files along with the model, or the folder, or the original .zip. The panel lists every image the file asks for with found or missing next to it. Matching ignores the folder part when it has to, so a texture sitting beside the .dae is found even if the file points at C:\Users\someone\textures.
Does it play Collada animations?
Yes, the first clip in the file plays in a loop, with a pause button. Skeletal animation and animated transforms both work. If the file is rigged but carries no animation, the model is shown in its bind pose and the panel says so.
Why does my SketchUp export look tiny or huge elsewhere but fine here?
Collada files declare their unit — SketchUp writes inches, for example — and many importers ignore it. This viewer applies it, so the size in the panel is in metres. It also applies the declared up axis, which is why Z-up SketchUp models stand upright without a switch.
Is DAE the same as Collada?
Yes. Collada is the name of the format; .dae (digital asset exchange) is its file extension. A .zae is a zipped Collada with its textures; rename it to .zip and drop it here.
Blender no longer opens my .dae. What now?
Blender dropped Collada support in 5.0. You can look at the file here, and to keep working with it convert it to GLB, which every current version of Blender imports.
Is my file uploaded?
No. The model and its textures are read from your disk and drawn in your browser. Missing textures are not fetched from anywhere either — the viewer never follows the paths in the file onto the network.
Can I make a VRM from a DAE character?
In two steps. Convert the .dae to .glb, then run the GLB to VRM converter, which maps the skeleton to the VRM humanoid. It works when the character has a recognisable humanoid rig; a static prop has nothing to map.
Move a Collada model to a current format
The DAE to GLB converter packs the mesh, skeleton, animation and textures into one .glb. From there the GLB to VRM converter can make an avatar of a rigged humanoid, and the GLB viewer shows the result.