Understand the project
Practical overview
When it fits
Use MediaPipe when a project needs on-device machine learning for live or streaming media and requires either task-level APIs and models or lower-level graph-based customization.
Who it is for
- Developers building computer-vision features for mobile, desktop, web or edge applications.
- Teams integrating audio, text or vision machine-learning tasks into applications.
- Engineers creating custom graph-based inference pipelines for local media processing.
Documented uses
- Add pre-trained vision, text or audio task capabilities to an application.
- Customize models with application-specific data using Model Maker.
- Assemble local media-processing workflows from packets, graphs and calculators.
- Visualize, evaluate and benchmark solutions in MediaPipe Studio.
- Build live or streaming media features that process input on supported devices.
Before you choose
Requirements & limits
The project is publicly available as open-source software under the Apache-2.0 license.
- MediaPipe Solutions offers task APIs, libraries and pre-trained models for vision, text and audio workflows.
- MediaPipe Framework provides lower-level packets, graphs and calculators for custom pipelines.
- The documented platform list includes Android, iOS, Linux, macOS and Windows, with self-hosted use also identified.
- MediaPipe Tasks process input on device according to the project documentation, while API performance and utilization metrics are sent to Google.
- The repository is implemented primarily in C++ and is described as cross-platform and customizable.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Evidence and transparency
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 1 October 2026
- Original website title
- google-ai-edge/mediapipe · Official homepage title · retained in its source language