How long is a Lyria 3.5 song, and how do I set the length?
Lyria 3.5 makes songs of a couple of minutes, and length is set in the prompt. What Google says, what Sume rejects, and prompts that steer duration.

A Lyria 3.5 song runs a couple of minutes, and you set the length in the prompt. Google's developer page says Lyria 3.5 produces "a couple of minutes" with duration controllable using the prompt. Sume's docs say the same: full-length structured songs up to a few minutes, steered by wording like "a 2-minute track", with no duration field.
Google's facts are from its Lyria page and Flow Music post; Sume's from Music 1.0 and the Music Router. Read 2026-09-29.
How do I ask for a specific length?
Put the length in words, or lay out time-stamped sections. Sume's docs give two forms:
- A phrase: "a 2-minute track" or "A 30-second track."
- Section markers with times:
[0:00-0:30] Intro: .... - A named moment in the brief: "breakdown to bass and claps at 0:20, full return at 0:28".
curl -X POST https://api.sume.com/v1/music-router/generate \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: length-45s-001" \
-d '{
"model": "lyria-3.5",
"prompt": "Uplifting corporate pop, 110 BPM, C major. [0:00-0:15] Intro: soft piano and pulse. [0:15-0:45] Full band with clean drums. A 45-second track. Instrumental, no vocals."
}'What happens if I send duration?
Sume rejects it. The Music Router and Music 1.0 both list duration and duration_seconds as unrecognized and rejected. Send only the prompt.
Will it match the length I asked for?
Treat the length as a target, not a promise. Sume's docs call prompt directions creative guidance and say to verify the generated audio. Google adds that results can vary between calls with the same prompt. Check the artifact's duration, and if the track must land on an exact length, trim it.
How do I cut a track to an exact length?
Use Timeline audio split on the Sume-hosted track with a ranges[] of { start, end }. It returns each range as its own file, WAV by default, for a flat per-job rate listed in the docs.
| Step | Detail |
|---|---|
| Generate | POST /v1/music-router/generate, length in the prompt |
| Trim | POST /v1/timeline-1.0/audio with operation: split |
| Input | A media.sume.com file; off-host URLs are rejected |
| Output | WAV by default, or MP3 with output.format |
Sources
Related posts
More in Models
- 2K AI video generator API: MiniMax H3 at 2K on Sume
MiniMax H3 makes video up to 2K. On Sume you ask for resolution 2K on minimax-h3, an upscale of native 768p. Request, price for 10 seconds, and limits.
- MiniMax H3 Max 1080p: a latent refinement from native 768p
MiniMax H3 Max on Sume offers 480p, 768p and 1080p. The 1080p tier is a latent refinement from native 768p. What that means for price and output checks.
- What is MiniMax H3 Max? The post-trained variant, explained
MiniMax H3 Max is a variant post-trained by fal.ai on MiniMax H3 for faster generation. Its resolutions, lengths and modes, and how Sume lists it.
- MiniMax H3 Max lip sync API: a still plus audio, 5 to 14.8 seconds
Sume runs MiniMax H3 Max lip sync at POST /v1/minimax/h3-max/lip-sync: send a still and Sume-hosted audio of 5 to 14.8 seconds. Body, resolutions and price.
Written by Sume