The Telegram bot for AudD, a real audio-recognition API company: send a voice message, audio file or video and it identifies the song playing in it.
AudD (@AudDBot) is the demo bot for AudD.io, a commercial music-recognition API used by other developers to add "what song is this" functionality to their own apps. Its own bio states it plainly: it "identifies music in voice messages, audio files, and videos," which covers a wider range of input types than most Shazam-style bots that only accept a clean recording.
Because it sits on top of a real recognition API rather than being a hobby project, the underlying matching technology is the same one paying developers rely on, which is a meaningfully different trust signal than an anonymous clone with no described backend. In practice you send audio — a voice memo of a song playing nearby, a video with music in the background, a straight audio file — and it replies with the track and artist if it finds a match.
It will not recognise music that is not in AudD's reference database (very obscure, unreleased or extremely regional tracks can miss), and like any audio-fingerprinting service it works best with a few clear seconds rather than a noisy, distant recording.
/startShow the welcome messageFree
| Criterion | Weight | Points | |
|---|---|---|---|
| Functionality | 25 | 8 / 10 | |
| Reliability & speed | 15 | 7 / 10 | |
| Popularity & trust | 15 | 5 / 10 | |
| Price & free tier | 10 | 8 / 10 | |
| Usability | 10 | 7 / 10 | |
| Safety & privacy | 10 | 7 / 10 | |
| Freshness | 5 | 6 / 10 | |
| Localization | 5 | 10 / 10 | |
| No intrusive ads | 5 | 6 / 10 | |
| Score | 100 | 70.5 |
Voice messages, audio files and videos with music in them, per the bot's own description.
It runs on AudD's commercial recognition API, the same technology paying developers use, though no independent accuracy figures are published.