GET

/v1/schemas/fal/video/fal-ai/sadtalker/reference

Accept:

kind inputactivity videohash 411ae982…extracted Jul 6, 2026version currentsource upstream spec ↗

field	recorded source
schema	Source linked above. Source trace `d703b4356edd8b147d5f493ee392c1a500fc6c030d3ad2628a2c3748f84978e7` { "fetchedAt": 1783331098, "extractorVersion": "1", "verifiedAt": null }

Request body

9 properties · 3 required · 0 $defs

property	type	description	constraints
driven_audio_url*	string	URL of the driven audio
source_image_url*	string	URL of the source image
reference_pose_video_url*	string	URL of the reference video
face_enhancer	const	The type of face enhancer to use	"gfpgan" · nullable
pose_style	integer	The style of the pose	0 ≤ n ≤ 45 · default: 0
face_model_resolution	enum	The resolution of the face model	256 \| 512 · default: "256"
still_mode	boolean	Whether to use still mode. Fewer head motion, works with preprocess `full`.	default: false
preprocess	enum	The type of preprocessing to use	crop \| extcrop \| resize \| full \| extfull · default: "crop"
expression_scale	number	The scale of the expression	0 ≤ n ≤ 3 · default: 1

Validate a payload

before you spend tokens

shell

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