Run Depth Anything V2 locally with OpenEye

    Estimate depth from a single camera. Enables spatial reasoning and 3D scene understanding without stereo cameras. OpenEye, the open-source perception CLI, downloads Depth Anything V2, runs it on your CPU or GPU, and returns structured JSON you can pipe into scripts, agents, and robots.

    TaskDepth estimation
    CLI iddepth-anything
    CreatorHKU / TikTok
    Weightsdepth-anything/Depth-Anything-V2-Small-hf (~98 MB)
    HardwareCPU, NVIDIA CUDA, Apple Silicon (MPS)
    Install extraopeneye-sh[depth]

    1. Install

    pipx install "openeye-sh[depth]"

    2. Pull Depth Anything V2

    openeye pull depth-anything

    3. Run inference

    Output is JSON on stdout, so it composes with jq and any other Unix tool.

    openeye run depth-anything scene.png --visualize

    4. Serve it as an API

    Start a local server with REST and WebSocket endpoints for apps, agents, and live camera streams.

    openeye serve depth-anything

    See the serve reference, REST endpoints, and Python SDK.

    Other models