Name: depth anything v2 small
Author: onnx-community

Overview

Depth Anything v2 (Small) is a lightweight monocular depth estimation model designed for real-time spatial analysis. Unlike traditional depth models that struggle with inconsistent scaling, v2 leverages a more robust training pipeline to provide highly precise relative depth maps from a single RGB image. For developers, the 'small' variant is optimized for edge deployment and low-latency applications where GPU memory is limited. It is particularly effective for robotics, AR/VR occlusion handling, and background blurring in video processing. Provided in ONNX format, it integrates seamlessly into cross-platform pipelines without requiring heavy framework dependencies, offering a superior balance of inference speed and depth granularity compared to its predecessor.

Highlights

High-precision monocular depth estimation in real-time
Optimized for edge devices and low-latency inference
ONNX format ensures easy cross-platform integration
Superior relative depth consistency over version one
Apache-2.0 license for flexible commercial deployment

Usage

Install

# Install Hugging Face transformers
pip install transformers torch

SDK Usage

# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("onnx-community/depth-anything-v2-small")
tokenizer = AutoTokenizer.from_pretrained("onnx-community/depth-anything-v2-small")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance

pip install -U huggingface_hub

CLI Download

Download the full repository

huggingface-cli download onnx-community/depth-anything-v2-small

Download a single file to a local folder (e.g. config.json into ./dir)

huggingface-cli download onnx-community/depth-anything-v2-small config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('onnx-community/depth-anything-v2-small')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://huggingface.co/onnx-community/depth-anything-v2-small

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/onnx-community/depth-anything-v2-small

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

pip install -U transformers torch

Load the model and run inference

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('onnx-community/depth-anything-v2-small')
tokenizer = AutoTokenizer.from_pretrained('onnx-community/depth-anything-v2-small')

Model Download

We recommend downloading the model via the ModelScope CLI or SDK.

Guidance：Before downloading, install ModelScope with:

Guidance

pip install modelscope

CLI Download

Download the full repository

modelscope download --model onnx-community/depth-anything-v2-small

Download a single file to a local folder (e.g. README.md into ./dir)

modelscope download --model onnx-community/depth-anything-v2-small README.md --local_dir ./dir

See the docs for more CLI options

SDK Download

# 模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('onnx-community/depth-anything-v2-small')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://www.modelscope.cn/onnx-community/depth-anything-v2-small.git

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/onnx-community/depth-anything-v2-small.git

ModelScope 模型页直接下载模型文件；无需将模型文件放在本站服务器。

Notebook Quickstart

Install the ModelScope library

pip install "modelscope[audio,cv,nlp,multi-modal,science]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html

Load the model and run inference

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

p = pipeline('text-generation', 'onnx-community/depth-anything-v2-small')

Full Documentation

来源: HuggingFace

---
base_model: depth-anything/Depth-Anything-V2-Small
library_name: transformers.js
license: apache-2.0
pipeline_tag: depth-estimation
new_version: onnx-community/depth-anything-v2-small-ONNX
---

https://huggingface.co/depth-anything/Depth-Anything-V2-Small with ONNX weights to be compatible with Transformers.js.

Usage (Transformers.js)

If you haven't already, you can install the Transformers.js JavaScript library from NPM using:

bash

npm i @huggingface/transformers

Example: Depth estimation w/ onnx-community/depth-anything-v2-small.

js

import { pipeline } from '@huggingface/transformers';
// Create depth estimation pipeline
const depth_estimator = await pipeline('depth-estimation', 'onnx-community/depth-anything-v2-small');
// Predict depth of an image
const url = 'https://huggingface.co/datasets/Xenova/transformers.js-docs/resolve/main/cats.jpg';
const { depth } = await depth_estimator(url);
// Visualize the output
depth.save('depth.png');

!image/png

---

Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using 🤗 Optimum and structuring your repo like this one (with ONNX weights located in a subfolder named onnx).