Music Generations

> Minimum-discount routing: Prefix the path with a min{N} segment (e.g. /min30/v1/music/generations) to require marketplace seller offers to meet a minimum estimated buyer discount before routing. Buyer-owned providers are not covered. See Minimum-Discount Routing.

Async music and text-to-audio generation via the marketplace. Same flow as video: submit, poll, retrieve.

Submit Job

POST /v1/music/generations

Auth: API key required (Authorization: Bearer *).

Request Body:

Field	Type	Required	Description
`model`	string	✅	Model ID, e.g. `venice-ace-step-15`; legacy `venice-music-gen` maps to ACE-Step
`prompt`	string	✅	Prompt or narration text (max 2000 chars)
`duration_seconds`	integer	Track/audio duration. Optional; omitted values use the model's default duration when known
`lyrics_prompt`	string	Lyrics text (max 5000 chars) for models that support lyrics
`force_instrumental`	boolean	Generate without vocals when supported
`voice`	string	Voice selection for TTS-like audio models
`language_code`	string	Language for lyrics or speech models when supported
`webhook_url`	string	URL for terminal state callbacks
`metadata`	object	Client pass-through metadata

Field Type Required Description
model string ✅ Model ID, e.g. venice-ace-step-15; legacy venice-music-gen maps to ACE-Step
prompt string ✅ Prompt or narration text (max 2000 chars)
duration_seconds integer Track/audio duration. Optional; omitted values use the model's default duration when known
lyrics_prompt string Lyrics text (max 5000 chars) for models that support lyrics
force_instrumental boolean Generate without vocals when supported
voice string Voice selection for TTS-like audio models
language_code string Language for lyrics or speech models when supported
webhook_url string URL for terminal state callbacks
metadata object Client pass-through metadata

Current Venice-backed examples include venice-ace-step-15, venice-elevenlabs-music, venice-minimax-music-v26, venice-stable-audio-25, venice-elevenlabs-sound-effects-v2, venice-mmaudio-v2-text-to-audio, and Venice audio TTS models. Query /v1/models and /api/markets for the live catalog and active order book.

Response: 202 Accepted

Same envelope as video:

{ "id": "550e8400-e29b-41d4-a716-446655440000", "object": "media.job", "kind": "music", "status": "queued", "created": 1713200000, "expires_at": 1713200900, "estimated_cost_usdc": "30000", "max_cost_usdc": "33000", "poll_url": "/v1/music/generations/550e8400-...", "cancel_url": "/v1/music/generations/550e8400-...", "job_token": "mjt_abc123..." }

Poll Status

GET /v1/music/generations/:id

Auth:** API key OR X-Job-Token header.

{ "id": "550e8400-...", "object": "media.job", "kind": "music", "status": "succeeded", "results": [ { "artifact_index": 0, "url": "https://media-artifacts.s3.amazonaws.com/async-media/job-id/0.mp3?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Expires=900&X-Amz-Signature=…", "download_url": "/v1/media/artifacts/job-id/0", "content_type": "audio/mpeg", "duration_seconds": 60, "bytes": 1200000 } ], "usage": { "seller_cost_usdc": "30000", "buyer_cost_usdc": "30000", "settlement_status": "confirmed" } }

Cancel Job

DELETE /v1/music/generations/:id

Same as video.

Differences from Video

Default expiry: 15 minutes.
Poll schedule is faster: 2s initial, 10s cap.
Supports lyrics_prompt, force_instrumental, voice, and language_code where the selected Venice model supports them.
Seller validation and offer health use /audio/quote so onboarding does not spend provider credits on full generations.
Download artifacts are cached/proxied for buyer access. Default artifact retention is 3 hours (MEDIA_ARTIFACT_TTL_SECONDS can override it).