Project AURIONAI Use-Cases Research & Innovation
vision

Depth Anything 3

Depth Anything 3 recovers depth and camera geometry from any number of views.

Depth AnythingFamily
Figure from the Depth Anything 3 paper: Given any number of images and optional camera poses, Depth Anything 3 reconstructs the visual space, producing consistent depth.
Levelvariant
Classvision
Parameters135.37M
Contextnot in the notes
Released13 Nov 2025
Opennessopen weights permissive
Runs onnot in the notes
Statusactive

Lineage

successor
Depth Anything V2 Large to Depth Anything 3: successor, high confidenceDepth Anything V2 LargeDepth Anything V2 LargeThe University of Hong Kong, Jun …Depth Anything 3Depth Anything 3Nov 2025
The University of Hong KongDepth Anything V2 Large is the parent (successor)high confidencenotes
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 2 models. Confidence and evidence are the notes' own.

R&D Topics

Monocular and stereo depth estimationproduced byFeed-forward 3D reconstruction and visual geometryproduced by

Identity

Developer
not in the notes
Family
Depth Anything
Released
13 Nov 2025 sourcesannouncement page
Weights published
–
Record checked
12 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
dense transformer
Total parameters
135.37M sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
image
Output
image
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
not in the notes
Footprint
not in the notes
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Benchmark Results

The notes record no benchmark result for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. arxiv.org/abs/2511.10647 https://arxiv.org/abs/2511.10647
  2. huggingface.co/depth-anything/DA3-BASE https://huggingface.co/depth-anything/DA3-BASE

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Depth Anything 3", https://aurion.semai.id/models/depth-anything-3/, records of October 2026.