google-ai-edge/mediapipe

Cross-platform, customizable ML solutions for live and streaming media.

View on GitHub

What it does

MediaPipe is Google's open-source toolkit that lets developers add AI-powered features — like face detection, hand tracking, and object recognition — to apps on phones, websites, and other devices without needing a connection to the cloud. It works in real time on video and camera feeds, meaning the AI processing happens directly on the user's device rather than being sent to a server.

Why it matters for PMs

With over 33,000 stars and broad platform support (Android, iOS, web), MediaPipe is one of the most widely adopted tools for building on-device AI features, which are increasingly table stakes in consumer apps — from fitness to AR to accessibility. For PMs and founders, this means teams can ship sophisticated AI-powered camera and media experiences faster and without the cost or privacy risks of cloud-based AI processing.

Early Signal Score25

Early stage — limited signal data

Stars
33.8k
Forks
5.8k
Contributors
97
Language
C++

Score updated Feb 18, 2026

Get the weekly digest

What just moved on gitfind.ai — delivered every Tuesday. No noise, just signal.