GET

/v1/schemas/fal/audio/fal-ai/vibevoice/7b

Accept:

kind inputactivity audiohash 861594de…extracted Oct 9, 2026version currentsource upstream spec ↗

field	recorded source
schema	Source linked above. upstream-spec Source trace `f4df8ff16fdb8ce77855e7f0e79e32a4c5ab02db2431f666497db479e2dfb8fd` { "fetchedAt": 1791522002, "extractorVersion": "1", "verifiedAt": null }

Request body

5 properties · 2 required · 1 $defs

property	type	description	constraints
speakers*	VibeVoiceSpeaker[]	List of speakers to use for the script. If not provided, will be inferred from the script or voice samples.
script*	string	The script to convert to speech. Can be formatted with 'Speaker X:' prefixes for multi-speaker dialogues.	length ≤ 30000
cfg_scale	number	CFG (Classifier-Free Guidance) scale for generation. Higher values increase adherence to text.	1 ≤ n ≤ 2 · default: 1.3
seed	integer	Random seed for reproducible generation.	nullable
enable_safety_checker	boolean	Authorized callers may set false to request relaxed screening.	default: true

Types

1 $defs

VibeVoiceSpeakerobject

preset	enum	Default voice preset to use for the speaker. Not used if `audio_url` is provided.	Alice [EN] \| Carter [EN] \| Frank [EN] \| Mary [EN] (Background Music) \| Maya [EN] \| Anchen [ZH] (Background Music) · +2 more · default: "Alice [EN]"
audio_url	string	URL to a voice sample audio file. If provided, `preset` will be ignored.	nullable

Validate a payload

before you spend tokens

shell

$ curl -X POST https://modelschemas.com/v1/validate \
    -d '{"provider":"fal","endpointId":"fal-ai/vibevoice/7b","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}