◎ Audio catalog API · v1.0.0

Recognize your audio.
Not the whole internet.

Upload the songs you care about. GrooveID fingerprints them, listens for snippets in any audio stream, and tells you what matched — in milliseconds.

All systems nominal Private libraries 1,000 free calls / mo
grooveid-quickstart.js
node
import { GrooveID } from "grooveid-sdk";

const client = new GrooveID({
  apiKey: process.env.GROOVEID_API_KEY
});

// Create a library namespace
const lib = await client.libraries.create("FM Radio");

// Upload your catalog
await client.songs.upload(lib.id, "./track.mp3");

// Recognize a live snippet
const result = await client.recognize(
  lib.id, "./clip.wav"
);
// → { matchFound: true, confidence: 94 }
FINGERPRINT SPECTRAL · 45S · 2048FFT SHA-256 ✓ PRIVATE NAMESPACES
PRIVATE CATALOG MATCHING · LOW LATENCY · API-FIRST
Built for developers

Six shipped use-cases. Probably yours too.

01

Broadcast monitoring

Detect when your licensed tracks air on radio or TV. Upload your catalog once, run continuous recognition against live streams.

02

DJ & live sets

Identify which track is playing in real time. Trigger lighting cues, BPM sync, or metadata display automatically.

03

Karaoke & education

Match a hummed melody or mic recording to a song in your library. Start lyrics sync or scoring when the right track is detected.

04

In-game triggers

Fire in-game events when a player performs or plays a specific song. Works with any audio input in real time.

05

Content moderation

Screen user-uploaded audio against a library of protected or flagged tracks before publishing.

06

Hum-to-find search

Let users hum or record a snippet to find a song in your platform’s catalog. No manual tagging needed.

How it works

From upload to match in four steps.

01

Get an API key

Sign up for a free account. Create an API key from the developer portal in under a minute.

02

Create a library

Libraries are private namespaces. Create one per project, client, or catalog segment.

03

Upload songs

Upload MP3, WAV, or M4A files. The API extracts a spectral fingerprint and discards the audio.

04

Recognize snippets

POST any audio clip. Get back a match result, confidence score, and metadata in milliseconds.

Your catalog Your match Your latency Your catalog Your match Your latency Your catalog Your match Your latency Your catalog Your match Your latency
FAQ

Answers to the first five.

Does it work on my own recordings?
Yes — that’s the point. GrooveID only recognizes what you upload. Your catalog stays private, separate from every other customer’s library.
What audio formats do you accept?
Song uploads accept common audio files including MP3, WAV, M4A, OGG, FLAC, AAC, WebM, and AIFF up to 10 MB. Recognition snippets use the same audio types up to 5 MB.
What makes a good snippet?
Use a clip from the part of the song that was fingerprinted. Five to fifteen seconds is ideal; clips under about one second are too short for reliable analysis.
Do you store the uploaded audio?
No. We extract a spectral fingerprint on ingest and discard the original file immediately. Your audio never touches persistent storage.
What’s the latency?
Every recognition response includes a latencyMs value. Actual latency depends on snippet length, library size, server load, and audio decoding time.
Ship it today.

Free tier, no card, 1,000 recognitions on us. If you need more, paid tiers are coming later this year.