sume/music-auto or pinned lyria-3.5: what changes in the job
Music Router takes sume/music-auto (Lyria 3.5 today), lyria-3.5 or lyria-3-pro. What job.model and routed_model echo back, and why the price stays the same.

Use sume/music-auto unless you need one engine. It is the default when model is omitted and resolves to Lyria 3.5 today. Pinning lyria-3.5 or lyria-3-pro passes through to that engine. Either way the price is the fixed Music price per audio generation, so the choice is about control, not cost.
The routable ids are sume/music-auto, lyria-3.5 and lyria-3-pro, per Music Router. Unknown ids fail with 400 model_not_found and a catalog_url.
What the job echoes
The model you requested is kept verbatim on the job; the engine that actually ran is recorded separately.
| You send | `job.model` | `job.request.routed_model` |
|---|---|---|
no model | sume/music-auto | the catalog engine, for example lyria-3.5 |
sume/music-auto | sume/music-auto | the catalog engine, for example lyria-3.5 |
lyria-3.5 | lyria-3.5 | lyria-3.5 |
Why log routed_model
Auto can change which engine it picks when the catalog changes. If you ship a track and later need to know what made it, routed_model is the field to store next to the artifact URL. Reading it costs nothing; it is on the job you already poll.
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-auto-001" \
-d '{
"model": "sume/music-auto",
"prompt": "Bright synth-pop, 118 BPM, A major. Gated snare, glassy arpeggio. A 30-second track. Instrumental, no vocals."
}'Old routes still work
The retirement of Music 1.0 is gradual. Its routes keep working and keep job.model = sume/music-1.0, but every request now resolves through the router, so routed_model is also present there. New integrations should call POST /v1/music-router/generate.
The price is $0.125 per generation on the Music 1.0 page, and the router page says every router model charges that fixed Music price. List the catalog with GET /v1/music-router/models for the per-model provider list prices.
Length is a prompt matter
Music Router rejects duration and duration_seconds; you ask for length in the prompt, for example "a 2-minute track" or a section map such as [0:00-0:30] Intro. The prompt can be 1 to 5000 characters, and negative_prompt must be omitted or empty, so put exclusions in the positive text ("Instrumental, no vocals").
Sources
Related posts
More in Models
- Does the Sume spend cap cover the agent's own model tokens?
No: Sume's cap excludes the agent's LLM turn; debited_usd_micros includes it. At list rates that turn is about $0.004 on Haiku 5.5 and $0.40 on Astra.
- Swap four people in one video: H3 Max Recast bills per second
h3-max-recast takes one source video and 1 to 4 person photos, 5 to 30 seconds. Price is $0.375 or $0.5625 a second whether you swap one person or four.
- Hy Image 3.5 Preview is not in Sume's catalog: what to call instead
Sume's image catalog does not list Hy Image 3.5 Preview. Check with one jq call, then pick a listed model by price: Soul, gpt-image-2.5, Seedream or Flux.
- Text-in-image bake-off: 20 prompts on 4 Sume models costs $7.00
Test text rendering in AI images for $7.00: 20 prompts on Ideogram 4.5, GPT Image 2.5, Qwen Image Max and Nano Banana 2.1, with the cost of each on Sume.
Written by Sume