Depth Anything 3 Base + Small - backbone + camera head (ONNX) for Perspectra
Redistribution of an existing open model, converted for the Perspectra Blender add-on. No retraining.
- Original work: ByteDance Seed, Depth Anything 3 (https://github.com/ByteDance-Seed/Depth-Anything-3), weights depth-anything/DA3-BASE and DA3-SMALL (Apache-2.0; the Large / Giant any-view models are CC BY-NC and NOT used)
- Licence: APACHE-2.0 — the same as the original weights. All credit to the original authors.
- Paper: Lin et al., Depth Anything 3: Recovering the Visual Space from Any Views, 2025
- Only the backbone and the camera head (the lens from several frames together); da3_base_cam_pos.npz is the positional-embedding table the graph takes as an input.
| File | Size | SHA-256 |
|---|---|---|
da3_base_cam.onnx |
362 MB | 09087234a7a5ee1f… |
da3_base_cam_pos.npz |
128 MB | 1a698d9c10cd6dd4… |
da3_small_cam.onnx |
91 MB | b23aa7eced49e3f9… |
da3_small_cam_pos.npz |
64 MB | 0ab1083e126e40f9… |
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support