LONDON, United Kingdom: Merlin Bird ID downloads reached 42 million worldwide, including 12 million in 2026, as AI-powered identification apps attracted new birdwatchers.
Jessie Barry, a Cornell Lab of Ornithology program manager working on Merlin, provided the figures to the BBC. She said the app now offers identification support for about 11,000 species across its different tools.
Cornell launched Merlin in the United States in 2014 with photo-recognition models covering 285 common species. The app now combines photographs, recorded sounds and a step-by-step identification tool.
However, its identification modes do not cover the same number of birds. Cornell says Sound ID currently supports 2,066 species, while its Photo ID documentation lists more than 6,900.
Merlin trains its machine-learning models with photographs and recordings submitted to eBird and the Macaulay Library. Cornell told the BBC that users had generated 4.4 billion Sound ID results.
Read: Genes tell story of how birdsong and human speech are nearly identical
Market intelligence company Sensor Tower estimated that downloads of rival app Birda increased by 31% in 2026.
Barry said Cornell had limited demographic data but was receiving more feedback from younger adults who had started birdwatching.
Stockport software engineer Sam Lewis has also developed Lifer, an app that turns bird sightings into collectable cards. Their appearance and value change according to the recording or photograph quality and the species’ rarity.
The Global Biodiversity Information Facility lists Lifer as an endorsed data publisher. However, Lifer says a future update will allow consenting users to contribute qualifying sightings; the feature is not yet generally available.
Read: Startup Operation Bluebird Seeks to Cancel X’s ‘Twitter’ Trademarks
Cornell advises users to verify Merlin’s suggestions independently before submitting records to eBird because the system can misidentify species.
