Project AURIONAI Use-Cases Research & Innovation
The University of Hong Kong
visionfamily

Depth Anything

Monocular depth estimation models trained on large-scale unlabelled images, used in robotics, mapping and AR.

The University of Hong KongThe University of Hong KongChinaByteDanceByteDanceChina
Diagram of the Depth Anything training pipeline with labelled and unlabelled images, from the model card.
Levelfamily
Classvision
Parametersnot in the notes
Contextnot in the notes
Released19 Jan 2024
Opennessopen weights permissive
Runs onmobile or NPU
Statusactive

Lineage

202420252026Depth Anything V2 LargeDepth Anything V2 Large, Jun 2024Depth Anything V2 Small, Jun 2024Depth Anything 3Depth Anything 3, Nov 2025
Depth Anything 3 belongs to this familyvision13 Nov 2025135.37M
The University of Hong KongDepth Anything V2 Large belongs to this familyvision13 Jun 2024335.3M
The University of Hong KongDepth Anything V2 Small belongs to this familyvision13 Jun 202424.8M
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. Confidence and evidence are the notes' own.

R&D Topics

Monocular and stereo depth estimationexemplar

Identity

Developer
The University of Hong Kong, ByteDance
Family
not in the notes
Released
19 Jan 2024 sourcesannouncement page
Weights published
22 Jan 2024
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
– sources
Active parameters
–
Checkpoint parameters
–
Note
Depth Anything V1, V2 (Small, Base, Large, Giant; metric indoor and outdoor) and Depth Anything 3 (multi-view).
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
mobile or NPU
Footprint
Edge or toolkit builds listed in: Qualcomm AI Hub, Apple Core ML.
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/2401.10891 https://arxiv.org/abs/2401.10891
  2. huggingface.co/LiheYoung/depth-anything-small-hf https://huggingface.co/LiheYoung/depth-anything-small-hf
  3. aihub.qualcomm.com/models/depth_anything https://aihub.qualcomm.com/models/depth_anything
  4. github.com/huggingface/transformers/tree/main/src/transf... https://github.com/huggingface/transformers/tree/main/src/transformers/models/depth_anything
  5. catalog.ngc.nvidia.com/orgs/nvidia/teams/tao/models/nvdepthanythingv2 https://catalog.ngc.nvidia.com/orgs/nvidia/teams/tao/models/nvdepthanythingv2
  6. developer.apple.com/machine-learning/models https://developer.apple.com/machine-learning/models/
  7. huggingface.co/apple/coreml-depth-anything-v2-small https://huggingface.co/apple/coreml-depth-anything-v2-small
  8. arxiv.org/html/2401.10891 https://arxiv.org/html/2401.10891

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