Professional bioacoustics in your pocket
BirdNET Live is your companion for when you’re out and about — record the birds along your route while you walk.
BirdNET Live is your companion for recognizing bird sounds – anytime and anywhere. Thanks to the powerful BirdNET+ AI, the entire analysis runs directly on your device. No internet connection is required, and your audio recordings and detections stay completely on your smartphone.
With its various modes, BirdNET Live adapts to different use cases. You can spontaneously explore the wildlife around you, carry out structured surveys, let the app run in the background while walking, or use your smartphone as a scheduled recording unit for longer surveys. Audio files you have already recorded can also be analyzed directly on the device. The individual modes are presented in detail on the following pages.
BirdNET recognizes more than just birds
Although the focus is on recognizing bird sounds, BirdNET Live now identifies considerably more animal groups. The app currently supports 9,789 animal species and identifies their sounds completely offline. In addition to the 8,927 bird species, it can also recognize 268 mammals, 340 amphibians, and 254 insects. This makes BirdNET Live one of the most comprehensive offline-capable applications for automatic animal sound recognition.
Of the roughly 11,000 bird species known worldwide, BirdNET Live can currently recognize 8,927 species.
Background
BirdNET Live is the next generation of the BirdNET app. It is made possible by a new generation of optimized AI models that are small and efficient enough to run directly on modern smartphones. As a result, the app combines professional bioacoustic analysis with full offline functionality.
BirdNET Live is the result of close interdisciplinary collaboration within the RangerSound project, funded by the German Federal Environmental Foundation (DBU). The goal of the project was to develop an AI-powered application that meets the demands of practical nature conservation and works reliably even under real-world conditions.
To achieve this, conservation biologists from the University of Würzburg, rangers from the Bavarian Forest National Park, hardware experts from OekoFor, and AI researchers from Chemnitz University of Technology worked closely together. Experiences from day-to-day fieldwork fed directly into the development of the app.
The result is a robust and practice-oriented open-source application designed specifically for field use. Combined with an external microphone, the smartphone can, for example, be stowed safely in a backpack or a waterproof bag while BirdNET Live continuously records and analyzes animal sounds in the background. This turns the smartphone into a powerful mobile bioacoustics system for research, nature conservation, and citizen science.
Overview of the modes
Depending on your use case, various modes are available:
Live Mode – Recognize animal sounds in real time. A live spectrogram and continuous detections make this mode ideal for spontaneously exploring nature.
Live ModePoint Count – Carry out standardized, time-limited surveys at a fixed location. This mode is especially suited to scientific mapping and monitoring projects.
Point CountSurvey Mode – Document longer walks, field visits, or transects. The app records your route via GPS and links detections to their respective locations.
Survey ModeARU Mode (Autonomous Recording Unit) – Turn your smartphone into an autonomous recording device. Scheduled recordings, for example at sunrise or sunset, enable long-term and unattended surveys in the field.
ARU Mode (Autonomous Recording Unit)File Analysis – Analyze existing audio recordings directly on your device. Supported formats include WAV, FLAC, MP3, and OGG files.
Explore – Use the BirdNET Geo model to find out which animal species are expected at your current location and during the selected season. You can also see which species you have already recorded.
Sessions – Manage past sessions, review detections with spectrograms and audio playback, add metadata, and export the results in various formats.