MediaPipe
developers.google.com
- Free plan
- Open source
- Windows
- macOS
- Linux
- Web
- iPhone and iPad
- Android
- API
- Self-hosted
What MediaPipe is
MediaPipe provides libraries and tools for adding AI and machine learning tasks to applications, with ready-to-run models and APIs developers can customize. Vision options include face and hand landmarks, gesture recognition, image classification and segmentation, object detection, and pose landmarks. Text tasks include language detection, classification, embeddings, proofreading and summarization; an audio classification task is also listed. MediaPipe Tasks supports Android, Web/JavaScript, Python and iOS. Its cross-platform APIs are described as enabling inference with as few as five lines of code, and its on-device pipelines use CPU, GPU and TPU acceleration for real-time use cases. MediaPipe Studio lets developers try web samples with text, webcam input or uploaded images and adjust model settings. Model Maker supports transfer-learning retraining for compatible models, but it is deprecated, no longer actively maintained, and cannot change a model's original task. Solution API inputs are processed on-device and not sent to Google, while usage and performance metrics are sent to Google. The low-level Framework supports building on-device pipelines; source-build guidance covers Linux, macOS, Windows and Docker, with some configurations marked experimental.
Who it suits
It suits developers who want to add on-device AI tasks to apps across supported mobile, web and Python environments. Teams considering model retraining should account for Model Maker's deprecated status and task limits.
Good to know, and what to watch
- Ready-to-run models and customizable APIs
- Vision, text and audio tasks available
- CPU, GPU and TPU acceleration
- Solution inputs are processed on-device
- Model Maker is deprecated and unmaintained
- Model Maker cannot change a model's original task
- Usage and performance metrics are sent to Google
Plans
Every plan and priceMediaPipe
MediaPipe open-source package
Open source
Getting started
- Open the MediaPipe Solutions guide
- Choose MediaPipe Tasks for Android, Web/JavaScript, Python, or iOS
- Select a ready-to-run vision, text, or audio task
- Use the cross-platform API to run inference
- Try web samples in MediaPipe Studio with text, webcam input, or uploaded images
- Build the Framework from source on Linux, macOS, Windows, or Docker when needed
The verdict
MediaPipe offers a broad set of application tasks and on-device pipelines for developers. Review the metrics handling and deprecated Model Maker before choosing its customization path.
Questions people ask
Is MediaPipe free?
Yes. MediaPipe is open source and the listed MediaPipe plan costs nothing.
Which platforms does it support?
Listed platforms include Android, iOS, Linux, macOS, self-hosted environments, web, and Windows. MediaPipe Tasks specifically supports Android, Web/JavaScript, Python, and iOS.
Can I use custom models?
Model Maker can retrain compatible models with transfer learning, but it is deprecated and cannot change the original model task.
Are inputs sent to Google?
Solution API inputs such as images, video, and text are processed on-device and are not sent to Google. Usage and performance metrics are sent to Google.
What can MediaPipe Studio do?
It lets developers try web samples with text, webcam input, or uploaded images and adjust settings such as confidence thresholds.
Is legacy MediaPipe Solutions supported?
Support for listed Legacy Solutions ended on March 1, 2023, although repository code and prebuilt binaries remain available as-is.
What the maker publishes
- Acceleration
- MediaPipe Tasks describes optimized on-device ML pipelines with end-to-end acceleration on CPU, GPU, and TPU for real-time use cases. source
- API extensibility
- MediaPipe Tasks exposes cross-platform APIs, while the lower-level Framework uses packets, graphs and calculators to build custom pipelines. source
- Audio feature
- MediaPipe provides an audio-classification task for categorizing sounds. source
- Browser evaluation
- MediaPipe Studio lets developers try web samples with their own text, webcam input, or uploaded images and adjust model settings such as confidence thresholds. source
- Browser evaluation
- MediaPipe Studio lets users test on-device models and pipelines in a browser, adjust settings and use camera, image or text inputs. source
- Community support
- The web setup guide directs users to Google AI discussions and Stack Overflow for questions and GitHub for bugs and feature requests. source
- Company founding
- Google's official history says Google Inc. was established in August 1998. source
- Company headquarters
- LinkedIn lists Google's headquarters in Mountain View, California. source
- Company size
- LinkedIn lists Google as a public software-development company with 10,001+ employees. source
- Current Python release
- PyPI lists MediaPipe 1.1.1, released October 14, 2026. source
- Customization
- Model Maker uses transfer learning to retrain compatible models with a developer’s data, and the guide suggests aiming for about 100 samples per class. source
- Customization limits
- Model Maker is deprecated and no longer actively maintained; retraining cannot change the task the original model was built to perform. source
- Desktop and self-hosted use
- The installation guide documents building and running the framework from source on Linux, macOS, Windows, and Docker, with some operating-system configurations marked experimental. source
- Developer platforms
- MediaPipe Tasks supports Android, Web/JavaScript, Python, and iOS; the Solutions guide’s availability table lists supported tasks by Android, Web, Python, and iOS. source
- Documentation depth
- The official site provides task guides, platform setup guides, API references, examples, framework concepts and build-from-source material. source
- Framework
- The low-level MediaPipe Framework is used to build efficient on-device ML pipelines and provides concepts including packets, graphs, and calculators. source
- Generative AI API
- The catalog includes an LLM Inference API and other generative-AI tasks, with LLM Inference currently listed for web. source
- Hardware acceleration
- The web BaseOptions delegate can select CPU or GPU execution, with CPU as the default. source
- Independent technical assessment
- GitNux describes MediaPipe as producing general pose and landmark outputs for motion-estimation workflows rather than application-specific event statistics. source
- Independent use case
- A Japanese academic prototype used MediaPipe Hands for real-time ten-finger landmark tracking, handedness, key mapping and typing-status visualization. source
- Legacy support
- Support for listed MediaPipe Legacy Solutions ended on March 1, 2023, though their repository code and prebuilt binaries remain available as-is. source
- Legacy support limit
- Google ended support for listed legacy solutions on March 1, 2023; unsupported legacy binaries remain available only on an as-is basis. source
- License
- The repository and Python distribution identify the software as Apache License 2.0. source
- Limit: Android crash
- Good First Issue indexes an open Android HolisticLandmarker crash involving protobuf-javalite. source
Checked against the maker's pages on 21 Oct 2026.