Suno v6 retires earlier models: pin an engine id for ad music
Suno says previous models retire as v6 rolls out. If an ad needs the same bed next quarter, pin an explicit Music Router id on Sume instead of auto.

Suno's v6 announcement says previous models are being retired as v6 rolls out (read 2026-10-07), so a prompt that gave you a bed last spring may not run on the same model next quarter. On Sume, reduce that risk by sending an explicit Music Router id such as lyria-3.5 instead of sume/music-auto, and by storing the finished audio file rather than the prompt.
What Suno changed
Suno launched v6 on 2026-09-09 as three models: v6 and the more experimental v6-wild for Pro and Premier subscribers, and v6-mini for everyone. The post lists editing a single section or lyric in plain language, mashups, sampling and isolating parts of a track, and text, audio, image and video inputs. It does not mention an API.
| Model | Access | Note |
|---|---|---|
| v6 | Pro and Premier | Flagship |
| v6-wild | Pro and Premier | More experimental |
| v6-mini | Everyone | Faster, more efficient |
| Earlier models | Being retired | As v6 rolls out |
Why this matters for a recurring ad bed
A campaign often wants the same music across a refresh: a 15-second cut now, a 30-second cut in six weeks. If the model behind the prompt changes, the same words give a different track. Lyria on Sume is also single-turn with no seed, so even an unchanged model will vary between runs. The only repeatable asset is the file.
- Keep the audio artifact from the winning job. It is a Sume-hosted
media.sume.comURL inresult.artifacts[]. - Record
job.request.routed_modelnext to it so you know which engine produced it. - Pin the engine id in config.
sume/music-autoselects Lyria 3.5 today, but that is a statement about today.
How to pin on Sume
Routable ids are sume/music-auto (default), lyria-3.5 and lyria-3-pro. Read them from GET /v1/music-router/models. All of them bill the same fixed Music price per accepted generation, so pinning costs nothing extra.
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: q4-bed-v1" \
-d '{
"model": "lyria-3.5",
"prompt": "Upbeat acoustic pop, 112 BPM, G major. Claps and a bright ukulele. A 30-second track. Instrumental, no vocals."
}'What Sume does not do
Sume does not route to Suno, and its Music inputs are a text prompt of 1 to 5,000 characters plus an optional image URL. There is no section-edit call, so a changed chorus means a new generation with the revised brief. For the licensing questions that sit on top of this, read each vendor's terms before an ad ships.
Sources
Related posts
More in Use cases
- Swap the presenter per market: one ad, two Recast jobs on Sume
Recast keeps the motion, camera, cuts and sound of one source ad and replaces the people. How to make one version per market, the limits and the price.
- SynthID on every Gemini TTS clip: keep the original file for ad review
Google marks every Gemini 3.8 TTS clip with SynthID and C2PA credentials. What that means for ad voiceovers, and what Sume docs do and do not say.
- 10 static ad variants in one Sume image request: n and its cap
The Sume Image API takes n from 1 to 10, but each model has its own cap. Read it from the catalog; use 4:5 on Nano Banana or 1088 x 1360 on GPT for feed ads.
- Test the hardest 5 seconds first: trim, then Recast before the full 30
Before paying $11.25 to recast 30 seconds at 768p, trim the worst 5 seconds and recast just that for $1.895. What the test can and cannot tell you.
Written by Sume