Skip to content
Agentmart
Submit a tool

TF.js Pose Detection (MoveNet)

Tracks body keypoints from a webcam or video using MoveNet running through TensorFlow.js.

Artifact by TensorFlow.js team

IndexedApache-2.0Science & mathsAI-poweredNo sign-upRuns in your browser

Use it

Opens storage.googleapis.com in a new tab. Nothing to install.

You provide
Webcam or video
You get
Skeleton keypoints overlay
Account
No sign-up needed
AI
AI-powered; no API key needed
Your input
Processed in your browser, not uploaded (stated by the tool or its repo)
Price
Free to use
Made with
Open-source web app
Source
Open source
Link check
Answered on Oct 31, 2026

About

TF.js Pose Detection (MoveNet) is an open-source web app in the science math category (pose estimation). It opens in your browser, so there is nothing to install. You provide webcam or video and get skeleton keypoints overlay. Your input is processed on your device and not uploaded. It needs no account and no API key. It uses an AI model to do its job. The source code is public on GitHub (tensorflow/tfjs-models).

Category: Science & maths · pose estimation

No reviews yet. Used it? Sign in with an established GitHub account to rate it. How reviews work

More science & maths tools

See all science & maths tools →

Details

Auth
No auth
License
Apache-2.0
First seen
Oct 31, 2026
Last checked
Oct 31, 2026
Source
GitHub

More from TensorFlow.js team

Cite this page

TF.js Pose Detection (MoveNet): free online tool. Agent Mart. https://agentmart.dev/artifacts/tf-js-pose-detection-movenet (last updated 2026-10-31)