AI Pose Estimation for React Native
Add MediaPipe-powered body tracking, rep counting, and form feedback to your React Native app — for both iOS and Android — with a single integration.
// Install
npm install quickpose-react-native-pose-estimation
// Usage
import { QuickPoseCamera, QuickPoseFeature } from
'quickpose-react-native-pose-estimation';
<QuickPoseCamera
features={[
QuickPoseFeature.Fitness.RepCounter.SQUAT,
QuickPoseFeature.RangeOfMotion.HipFlexion,
QuickPoseFeature.Yoga.PoseDetection,
]}
onResults={(results) => handleResults(results)}
/>
Everything you need, out of the box
Pre-built, production-ready AI features you can drop into any React Native app — no ML knowledge required.
Rep Counter
Automatic rep counting for squats, push-ups, lunges, curls, and more. Anti-cheat detection flags partial reps — the ideal gating feature for a premium tier.
Pose Estimation
Real-time 33-point body landmark detection powered by Google MediaPipe. Track every joint position at 30fps directly on the device.
Range of Motion
Measure joint flexion angles for hip, shoulder, knee, and more. Used in rehab, physio, and health apps to track mobility and functional recovery.
Yoga Pose Detection
Recognise and score yoga poses in real time. Highlight skeleton overlays and give users live alignment coaching without a human instructor.
Sports Biomechanics
Frame-by-frame biomechanics analysis for athletic technique. Identify inefficiencies in a swing, jump, or sprint and surface feedback to users instantly.
Joint Overlay & Visualisation
Customisable skeleton overlays and joint highlights rendered directly on the camera view. Fully brandable — your UI, your design.
One plugin. Both platforms.
Write your AI fitness features once in React Native and ship to iOS and Android from a single codebase. No separate native integrations required.
iOS
Supports iPhone and iPad. Requires iOS 14.0+ and a physical device with a rear camera.
Swift Package ManagerAndroid
Supports Android phones and tablets. Requires Android SDK 26+ and a physical device with a rear camera.
Gradle / MavenBuilt for every movement app
QuickPose powers a wide range of fitness, health, and sports applications across iOS and Android.
Home Fitness
Transform a basic workout timer into an AI personal trainer with rep counting, anti-cheat detection, and real-time form feedback.
Yoga & Mindfulness
Guide users into correct alignment with live pose scoring and joint overlay coaching during on-demand or live sessions.
Sports Performance
Give athletes real-time biomechanics feedback on their technique — analyse movement frame-by-frame using live or recorded video.
Rehab & Physiotherapy
Track joint range of motion and exercise compliance between clinic visits. Clinical-grade accuracy, no specialist hardware needed.
On-device. No video leaves the phone.
QuickPose processes everything locally using the device's Neural Processing Unit. No video is ever uploaded to a server.
- No camera data sent to any server or API
- No per-session cloud costs — inference is free at any scale
- GDPR and CCPA compliant by architecture
- Works fully offline — no network connection required
- App Store & Google Play ready
From install to running in minutes
QuickPose integrates into any React Native app in a single afternoon — with no prior ML experience needed.
Install the package
Add QuickPose to your project via npm or Yarn.
npm install quickpose-react-native-pose-estimationLink native dependencies
Run pod install for iOS. Android links automatically via Gradle.
cd ios && pod installDrop in the camera view
Import QuickPoseCamera and choose your features — rep counter, pose detection, range of motion, and more.
Ship on both platforms
Build once, deploy to iOS App Store and Google Play from the same codebase. No separate native SDKs to maintain.
Common questions
QuickPose requires access to native camera APIs and the device's Neural Processing Unit. For this reason it requires a bare React Native workflow or an Expo bare workflow with custom native modules. It is not compatible with Expo Go out of the box.
Yes. QuickPose uses the device camera for real-time pose estimation, so a physical device is required. iOS simulators and Android emulators do not have camera access and cannot run QuickPose features.
iOS 14.0 or later and Android SDK 26 (Android 8.0) or later. Both platforms require a device with a rear-facing camera.
No. All pose estimation runs entirely on the user's device using the Neural Processing Unit. No video, image, or biometric data is ever transmitted to QuickPose servers or any third party. This makes GDPR compliance straightforward — there is no data transfer to disclose.
QuickPose is free to integrate and develop with, up to 10 monthly active devices. Paid plans start at $25/month for up to 100 devices. See our pricing page for full details.
The React Native plugin wraps both the iOS Swift SDK and Android Kotlin SDK, exposing the same set of features through a single JavaScript/TypeScript interface. If your team works in a native codebase, the platform-specific SDKs offer the deepest level of customisation. The React Native plugin is the fastest path to shipping on both platforms from one codebase.
