AI music BPM control: set the tempo in the prompt
There is no BPM parameter in Lyria 3.5 or Sume's music API. Write the tempo as a number in the prompt, then check the result. How, and what to expect.

You control the BPM of AI-generated music by writing it into the prompt, as a number: "72 BPM" or "142 BPM half-time". Neither Sume's music API nor Google's Lyria 3.5 documents a tempo parameter; Sume's docs say the request has no seed, temperature, guidance or duration field, and the tempo is a creative direction, not a guaranteed setting.
Sume's guidance is from the Music 1.0 and Music Router docs. Google's Lyria page says to mention instruments, BPM, key and mood, and its Flow Music post lists easier tempo control as a Lyria 3.5 gain. Read 2026-09-29.
How do I write the tempo?
Give a number and a feel, and pair it with a key and lead instruments, as the Music 1.0 brief does. One line of the docs' example: "Hushed and slightly melancholic neo-soul nocturne, 72 BPM, D minor."
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: bpm-128-001" \
-d '{
"prompt": "Driving synthwave, 128 BPM, F minor. Pulsing bass, gated snare, bright arpeggio. A 1-minute track. Instrumental, no vocals."
}'Will the track hit that exact BPM?
Not guaranteed. Sume's docs call the brief axes "creative directions, not guaranteed output settings" and tell you to verify the generated audio. They also note that a lyrics field, when present, is model-reported metadata and not an audio measurement, so do not read tempo off it. Measure the track in an audio editor or a BPM tool and regenerate if it is off.
How do I keep several tracks in the same tempo?
Repeat the number and the key in each brief, and change the arc. For contrasting scenes, the docs go the other way: vary the tempo by at least 12 BPM, the genre family and the lead instrument.
| Goal | What the docs say to do |
|---|---|
| One consistent score | Keep the brief axes close; pass the accepted scene still as image_url when it fits |
| Contrasting scenes | Vary genre family, tempo (at least 12 BPM apart) and lead instrument |
| Tempo as a number | Write it in the prompt, e.g. "72 BPM" |
| Length | Say it in the prompt; duration is rejected |
Can I change the tempo after the track exists?
Not through Sume's music API, and Google says Lyria music generation is single-turn: refining a result with more prompts is not supported. Generate a new take with a changed number in the prompt. Each take is billed as its own generation.
Sources
Related posts
More in Media tools
- AI music stems: does the API return separate tracks?
Sume's music API returns one audio file, not stems. What Lyria 3.5's documentation covers, and how to get parts you can mix separately.
- AI song structure tags: [Verse], [Chorus] and timestamps
How to structure an AI song prompt for Lyria 3.5 with section tags and timestamps, and how the same prompt goes through Sume's Music Router.
- AI sound effects for video: what Sume lists and what it does not
Sume does not list a dedicated sound-effects model. Sound for video comes from video models with generate_audio, the music route, or your files on the Timeline.
- Amazon online video ad specs: OLV size, length, bitrate
Amazon online video (OLV) ads run 6–120 s in 16:9, at least 1920×1080 and 4 Mbps, with 192 kbps AAC on 2+ channels and up to 500 MB site-served.
Written by Sume