Lyria 3.5 in another language: prompt in it, then check the song
Google says Lyria 3.5 makes music in other languages when you prompt in that language. On Sume, write the brief and lyrics in it, then check the result.

To get a Lyria 3.5 song in another language, write the prompt and any lyrics in that language. Google's docs say that is how multilingual music is requested (read 2026-10-02). On Sume the same applies through Music Router, where sume/music-auto runs Lyria 3.5 today.
What Google documents
The Gemini API music page lists lyria-3.5 as the full-length model, with a couple of minutes of music controllable from the prompt, and says lyrics and song structure come back alongside the audio (read 2026-10-02). It also notes that results may vary between calls even with the same prompt, that generation is single-turn, and that all audio carries a SynthID watermark (read 2026-10-02).
| Item | Google says |
|---|---|
| Model id | lyria-3.5 |
| Length | Full songs, about a couple of minutes, steered by prompt |
| Output | MP3 by default, WAV for Lyria 3.5; 44.1 kHz stereo |
| Language | Prompt in the target language |
| Repeatability | Results can differ between calls |
How to do it on Sume
Send a normal Music Router request with the language named in the prompt and the sung text written in that language. Use section tags or timestamps to give structure, for example [0:00-0:30] Intro:, which the Sume docs list as a way to steer length and layout.
Read result.lyrics after the job completes. The docs describe it as model-reported lyrics or a section map when present, so use it to confirm the words came back in the right language before you caption the track.
Checks before you ship
A prompt in the target language is a request, not a guarantee. Listen for pronunciation and drift, and regenerate if needed; each accepted generation is billed at the fixed Music price.
Google's safety filters block specific artist voices and copyrighted lyrics (read 2026-10-02), so write original lyrics rather than quoting a known song.
Sume's docs do not publish a language list for music, so test the language you need rather than assuming support.
Sources
Related posts
More in Models
- MAI-Image-2.6 output cap is 2,359,296 pixels; Sume sizes differ
MAI-Image-2.6 sets a 2,359,296-pixel ceiling and a 768-pixel minimum edge. Sume sets size per model with tiers, ratios and, for GPT models, custom pixels.
- MAI-Image-2.6 edits take 5 references; Sume takes 10 or 16
MAI-Image-2.6 in Foundry accepts up to five JPEG or PNG reference images per edit. On Sume, input_references tops out at 10, or 16 on GPT Image 2.5.
- MAI-Image-2.6 web_grounding flag: what Sume has instead
MAI-Image-2.6 can pull Bing results into an image when web_grounding is on. Sume has no such flag; here is how to pass current facts in the prompt instead.
- Meta Muse Video: preview only, so what can you use for Reels?
Meta's July 7, 2026 post previews Muse Video with native audio and says it is coming soon. No API details yet. How to check what Sume lists for 9:16 video.
Written by Sume