Shotcraft
Music and sound

Google Lyria 3 Clip

The short-form Lyria 3 variant, model ID lyria-3-clip-preview, which produces a fixed 30-second clip from a text prompt with optional lyrics and image references.

An audio model by Google · read 2026-08-14 · sourcing high

Up to
30 seconds
Resolution
44.1 kHz stereo MP3
Weights
Not released
API
Yes
Commercial use
Permitted subject to conditions

Intended for loops, previews, social assets and rapid prototyping rather than full songs. Output is 44.1 kHz stereo MP3 with a SynthID watermark.

Commercial use — Permitted subject to conditions

Same terms basis as Lyria 3 Pro: no output-ownership clause is published on the model documentation; use is governed by the Google Cloud / Gemini API terms and the Generative AI Prohibited Use Policy (https://policies.google.com/terms/generative-ai/use-policy).

Who owns the output: depends on your plan tier

  • Depends on your plan tier. Per-tier commercial rights and indemnification not verified from a terms page
  • The terms restrict what you may generate. Same terms basis as Lyria 3 Pro: no output-ownership clause is published on the model documentation; use is governed by the Google Cloud / Gemini API terms and the Generative AI Prohibited Use Policy (https://policies.google.com/terms/generative-ai/use-policy)
  • Output is watermarked. All output is SynthID-watermarked and prompt-level safety filters block requests for specific artists' voices or copyrighted works

The conditions above were extracted from the licence notes recorded during research, and each one shows the sentence it came from so you can check the reading. Read the licence before you rely on any of this. Nothing on this page is legal advice, and this site does not tell you a model is cleared for commercial use.

Specification

Vendor figures, read 2026-08-14Read from Google's own documentation on 2026-08-14. Vendors change these between model versions without announcing it, so check theirs before you commit a shot to a number. Sourcing for this entry is strong: the figures came from pages we retrieved and read. Check the vendor's current documentation

What it takes in
text, lyrics, image
What it makes
audio
Maximum duration
30 seconds
Maximum resolution
44.1 kHz stereo MP3
Controls
lyrics-input, style-prompting, instrumental-only
Weights
Not released
API
Yes
Pricing
$0.04 per song for 'Lyria 3 Clip Preview (30s)' on the Gemini API, no free tier.
Released or updated
2026-04-08

Strengths

  • Cheapest verified per-generation price in this set at $0.04/clip
  • Fast enough for bulk short-form asset generation
  • Same lyric and image conditioning as Lyria 3 Pro

Weaknesses

  • Hard 30-second cap, MP3 only
  • Preview status with restrictive rate limits
  • No stems, extend or editing
  • Watermarked output

Best for

Where the sources disagree

Recorded rather than resolved. Picking the tidier number would hide the fact that the vendor's own pages do not agree.

What to understand before using it

Explainers naming the mechanisms this model's own documentation mentions.

MCP servers that front Google

Servers in the media MCP directory whose own description or homepage names this vendor. Absence of a server here is not evidence that none exists: most publish no tool list at all.

See it from the vendor

Shotcraft hosts no sample output. Judging a model means watching its own showreel with your own eyes, and a clip re-hosted here would be a clip we chose.

Nearby in the directory

Google Lyria 3 Pro

Generates full songs with vocals and timed lyrics from a text prompt, optional user-supplied lyrics and up to ten reference images, at up to about thr

Music and sound

Google Lyria RealTime

An experimental streaming music model (models/lyria-realtime-exp) that generates continuous instrumental music over a bidirectional WebSocket while yo

Music and sound

ACE-Step

An open-source music generation foundation model combining diffusion with a lightweight linear transformer, released May 2025 by ACE Studio and StepFu

Music and sound

Sources

Read on 2026-08-14. A source that has changed since then will not be reflected here until the next pass.