Music Router: does pinning lyria-3-pro cost more than music-auto?
On Sume's Music Router every routable id, sume/music-auto, lyria-3.5 and lyria-3-pro, charges the same fixed Music price per generation. What pinning changes.

No. On Sume's Music Router, every routable model charges the same fixed Music price per audio generation, so pinning lyria-3-pro or lyria-3.5 costs the same as leaving model at sume/music-auto. The Music 1.0 page lists that price as $0.125 for each accepted generation, and says it does not change with prompt length or optional image conditioning. The router page states that the catalog lists the provider list price per model for reference only.
What pinning does change
The router has three routable ids: sume/music-auto (the default), lyria-3.5 and lyria-3-pro. If you omit model or send sume/music-auto, Sume selects the engine, which is Lyria 3.5 today. If you pin an id from GET /v1/music-router/models, the request passes through to that engine.
Two fields tell you what ran. job.model echoes the id you requested, so sume/music-auto stays sume/music-auto. job.request.routed_model names the catalog engine that ran, for example lyria-3.5. Log the second field if you ever compare tracks.
Pin or let Sume pick
Because the price is flat, an engine choice is a quality and consistency decision, not a budget decision. The only way to settle which is better for your brief is to render the same brief on each id and listen.
| You send | Engine | Price per generation | Use when |
|---|---|---|---|
| model omitted or sume/music-auto | Chosen by Sume (Lyria 3.5 today) | Fixed Music price | You want Sume to update the engine for you. |
| lyria-3.5 | Lyria 3.5 | Fixed Music price | You want a track to be repeatable by engine name. |
| lyria-3-pro | Lyria 3 Pro | Fixed Music price | You want that engine, and you accept the same price. |
| Any other id | None | None | 400 model_not_found with catalog_url. |
A request that pins an engine
The body is the same as Music 1.0, plus model. Control length in the prompt (for example A 30-second track). The API rejects duration and duration_seconds, and a non-empty negative_prompt.
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: music-router-pin-001" \
-d '{
"model": "lyria-3-pro",
"prompt": "Warm lo-fi hip hop, 84 BPM, C minor. Dusty Rhodes chords, brushed drums. A 30-second track. Instrumental, no vocals."
}'Before you rely on it
Music 1.0 is retiring gradually. Its routes keep working and keep job.model = sume/music-1.0, but every request resolves through the router. New integrations should call POST /v1/music-router/generate.
Read the live rate with GET /v1/catalog before a large batch. The docs name the flat price, and the catalog is the source for the current number.
Sources
Related posts
More in Pricing
- Nano Banana 2 edit chain cost on Sume: 1K, 2K and 4K per pass
A Nano Banana 2 edit costs $0.10 at 1K, $0.15 at 2K and $0.20 at 4K on Sume. Cost of 3-, 5- and 10-pass chains, and when to draft at 1K and finish at 4K.
- Nano Banana 2 thumbnail tier: Google 0.5K to 4K vs Sume billed
Google's 2026 Nano Banana 2 per-image prices from 0.5K to 4K next to what Sume bills, to pick a thumbnail tier by cost rather than habit.
- Nightly voice-agent test calls: 200 utterances, a month of TTS
A nightly regression run of 200 utterances at 100 characters is 600,000 characters a month: $9.00 on MAI Flash, $13.20 on MAI-Voice-2.1, $28.50 on Sume.
- 360p drafts before a 1080p final: when Omni drafting pays off
A 10 s Omni 360p draft costs $0.38 and a 1080p take $1.88. From the second attempt on, drafting first is cheaper. Break-even table on Sume, read 2026-10-05.
Written by Sume