Loading models…
Model Categories
Song Generation: Create full songs with optional lyrics and vocal support- ACE-Step 1.5, ElevenLabs Music, MiniMax Music 2.0
- Stable Audio 2.5
- ElevenLabs Sound Effects, MMAudio V2
Audio generation uses an async queue system. See the Audio Queue API to start generation and Audio Retrieve API to fetch results.
Pricing
Pricing varies by model:- Per-generation: Fixed price per audio clip (MiniMax Music 2.0, Stable Audio 2.5)
- Duration-tiered: Price scales with duration tier (ElevenLabs Music, ACE-Step 1.5)
- Per-second: Price based on output duration (ElevenLabs Sound Effects, MMAudio V2)
Duration-Tiered Pricing
Models with duration-tiered pricing accept anyduration_seconds within the model’s min_duration–max_duration range. The price is determined by which tier the requested duration falls into. Tier ranges are returned in the /models response under pricing.durations, with min_seconds and max_seconds for each tier.
For example, ElevenLabs Music accepts 3–600 seconds (up to 10 minutes) at $0.75 per minute, rounded up to the nearest minute:
| Duration Range | Tier Key | Base Price |
|---|---|---|
| 3–60s | 60 | $0.75 |
| 61–120s | 120 | $1.50 |
| 121–180s | 180 | $2.25 |
| 181–240s | 240 | $3.00 |
| 241–300s | 300 | $3.75 |
| 301–360s | 360 | $4.50 |
| 361–420s | 420 | $5.25 |
| 421–480s | 480 | $6.00 |
| 481–540s | 540 | $6.75 |
| 541–600s | 600 | $7.50 |
Key Parameters
| Parameter | Description |
|---|---|
prompt | Text description of the audio to generate |
lyrics_prompt | Song lyrics for vocal models (required when model has lyrics_required=true) |
duration_seconds | Output length in seconds |
force_instrumental | Generate without vocals (where supported) |