Py-Feat review
A research-focused toolkit for analyzing facial expressions in images and video locally.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
Py-Feat is a Python toolbox for researchers working with facial-expression data. It analyzes images and video for faces, expressions, Action Units, emotion categories, landmarks, gaze, and head pose. It also estimates continuous valence and arousal. The core package runs locally, while Py-Feat Live is a separate desktop app for real-time webcam analysis. That makes Py-Feat a fit for research workflows that need interpretable facial measures and local processing, rather than a hosted service or a general-purpose emotion AI platform.
Its two detector versions address different analysis needs. The modular pipeline lets users select or disable models; the multi-task model adds valence/arousal and 3D face-mesh outputs. Video-frame batching can support processing video sequences, and plotting and visualization utilities help inspect results. The toolbox is available for Windows, macOS, and Linux. These capabilities make it useful where researchers want to assemble image or video analyses around specific outputs. Buyers looking for a direct API integration should look elsewhere: the core package is installed as Python software, not exposed as an API.
Py-Feat is free and open source, with documentation as its listed support channel. The project license is MIT, but that does not necessarily cover every included model; some models carry non-commercial stipulations. Teams should check the terms for the specific models in their workflow before adopting it, particularly for commercial use. Researchers comfortable installing and working with Python can consider Py-Feat for local analysis and visualization. Those seeking a managed service, API access, or a uniform commercial license across all models should choose an alternative that explicitly provides those options.
Py-Feat pros and cons
- Where it wins
- Analyzes Action Units, emotion categories, valence, and arousal
- Includes gaze, head-pose, landmark, and 3D face-mesh outputs
- Modular models and video-frame batching support varied workflows
- Where it doesn't
- Requires a Python package installation for the core toolbox
- Core package has no API access
- Some included models have non-commercial stipulations
Py-Feat fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update