BirdNET App
Identify birds around you by recording a short snippet of song or call. Fast, free, and powered by BirdNET AI.
BirdNET app activity
Overview
The BirdNET mobile app lets you hold up your phone, record a few seconds of sound, and get instant identification suggestions. It uses the same core neural network models that power research-grade monitoring tools, simplified for everyday use.
- Real-time bird sound identification.
- Confidence scores and alternative suggestions.
- Optional contribution of observations for research.
- Works in many regions worldwide (model updates expand coverage).
Designed for birders, learners, educators, and anyone curious about the sounds they hear outdoors.
Representative interface screenshots.
Simple capture
Tap record, wait for the birds to call, select a clip, and submit for analysis. No external microphone needed.
Instant results
Top species suggestions with confidence help you learn local birds faster.
Learning tools
Link detections to species info and improve recognition through repetition.
How it works
1. Record
Phone microphone captures ambient sound (short clip).
2. Analyze
Clip converted to a spectrogram; BirdNET model scores species.
3. Suggest
Top matches returned with confidence and alternate options.
4. Learn
Users explore species info, improving recognition over time.
Data & privacy
Your recordings help improve BirdNET. Submissions are used for research (e.g., to train and evaluate models) and are licensed under CC BY‑SA 4.0.
- Human speech is filtered; we do not publish clips containing speech. Published examples are anonymized.
- Location is optional and user‑controlled; when provided, it helps track species diversity and change over time.
- Data are stored securely to prevent unauthorized access.
- Used for non‑commercial research and model development only.
See the in‑app policy screen for additional details.
Tips
- Move closer to the singing bird when possible.
- Avoid handling noise (hold device steady while recording).
- Limit overlapping sounds (traffic, voices) for better accuracy.
- Try multiple short recordings if the first is uncertain.
BirdNET Live (Web App)
Identify birds in real-time directly in your browser. BirdNET Live is a Progressive Web App (PWA) that processes sound locally on your device using TensorFlow.js.
- Run offline: No audio uploaded, works without internet.
- Real-time: Instant spectrogram & predictions.
- Cross-platform: Desktop & mobile browsers.
- Location-aware: Filters by local species.
Note: This project is in active development. Features and performance may vary.
Add to home screen for app-like experience.
Support & feedback
If you encounter issues or have improvement suggestions, please use the feedback option in the app or contact the team.
Email: ccb-birdnet@cornell.edu