GET

/v1/schemas/fal/audio/fal-ai/diffrhythm

Accept:

kind inputactivity audiohash 4592cf7c…extracted Jul 6, 2026version currentsource upstream spec ↗

field	recorded source
schema	Source linked above. Source trace `05fc4a6dd2970b37df37075272cd594a34aa7a8cc1a8f562116e76a164c86347` { "fetchedAt": 1783330412, "extractorVersion": "1", "verifiedAt": null }

Request body

7 properties · 1 required · 0 $defs

property	type	description	constraints
lyrics*	string	The prompt to generate the song from. Must have two sections. Sections start with either [chorus] or a [verse].
cfg_strength	number	The CFG strength to use for the music generation.	1 ≤ n ≤ 10 · default: 4
reference_audio_url	string	The URL of the reference audio to use for the music generation.
style_prompt	string	The style prompt to use for the music generation.
num_inference_steps	integer	The number of inference steps to use for the music generation.	10 ≤ n ≤ 100 · default: 32
music_duration	enum	The duration of the music to generate.	95s \| 285s · default: "95s"
scheduler	enum	The scheduler to use for the music generation.	euler \| midpoint \| rk4 \| implicit_adams · default: "euler"

Validate a payload

before you spend tokens

shell

$ curl -X POST https://modelschemas.com/v1/validate \
    -d '{"provider":"fal","endpointId":"fal-ai/diffrhythm","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}