lyrcs.ai API

Lyrics infrastructure for music at scale.

Submit audio. Receive transcription, transliteration, translation, time-synced LRC, SRT, and word-level JSON timestamps — via a single API call. Built for South Asian music. Ready for the world.

One API, three modes

Choose what you need.

Every job runs in one of three modes. Select based on what output you need and how fast.

align=false

Transcript only

Transcript + transliteration + translation + cultural notes. No LRC or SRT generated. For metadata delivery, streaming store submission, liner notes, and sync licensing.

~60s
align=truedefault

Full pipeline

Everything above, plus LRC and SRT files. For karaoke, lyric display apps, and time-synced lyrics on streaming platforms.

Includes word-level JSON timestamps by default. Set word_align: false for line-level only.

~90s
align=truereview=true

Reviewed delivery

Full pipeline output, held until the artist approves via a review URL. For artist approval workflows before public delivery.

~90s + review
POST /v1/align

Align only

You supply the lyrics. We return word-level timestamps — no transcription step. For catalogues that already have lyrics.

~45s

Simple to integrate

One endpoint. Async results.

POST https://lyrcs.ai/api/v1/transcribe
Authorization: Bearer lyrcs_sk_live_...
{
"audio_url": "https://cdn.example.com/songs/ek-do-teen.mp3",
"language": "Hindi",
"align": true,
"word_align": true,
"webhook_url": "https://api.yourdomain.com/webhooks/lyrcs"
}
→ 202 Accepted
{
"job_id": "a1b2c3d4-...",
"status": "queued",
"language": "Hindi"
}
// Webhook on completion:
{
"downloads": {
"lrc_original": "https://lyrcs.ai/api/v1/jobs/.../download/lrc/original",
"srt_original": "https://lyrcs.ai/api/v1/jobs/.../download/srt/original",
"words_original": "https://lyrcs.ai/api/v1/jobs/.../download/words/original"
}
}

Results delivered via webhook when ready. Poll GET /jobs/{id} for status. Read the docs →

Language coverage

48 languages. Deeper accuracy for South Asian music.

30 Indian languages including all 22 scheduled languages. Language-specific correction layers handle the phonological nuances that general models miss — aspirated consonants, retroflex sounds, nasal markers, script-specific sandhi rules.

HindiTamilTeluguPunjabiBengaliMarathiKannadaMalayalamGujaratiUrduand 38 more

Who it's for

Built for organisations that handle music at scale.

Music distributors

Deliver time-synced lyrics to DSPs alongside audio.

Music labels

Process catalog across artists and languages with optional artist approval.

Regional streaming platforms

Power lyric display without building transcription infrastructure.

Rights management companies

Structured lyric metadata for sync licensing and catalog databases.

Lyrics and metadata providers

Accurate, time-synced Indian language lyrics data at scale.

Karaoke providers

LRC files and word-level JSON timestamps from raw audio at scale. Apple Music-style karaoke for any language.

Apple Music & DSPs

Word-by-word karaoke timestamps for Apple Music Lyrics, TIDAL, and next-generation lyric display. The only API with native South Asian language support.

48
languages
~90s
full pipeline · word-level default
20
jobs per batch
HMAC-SHA256
webhook signing

Ready to integrate?

Tell us about your catalog and we'll set you up with API access within 2 business days.

Or read the docs first.