Create Higgsfield Speak v2 lip-synced talking video job

Queues a paid lip-sync request that animates a creator image speaking the supplied script. The script must stay within 40 words (about 15 seconds of audio); BAMF picks the smallest 5, 10, or 15 second Speak v2 duration that fits. Product-launch founder shots select Higgsfield Speak explicitly; failures do not silently forward likeness or cloned-voice data to another video vendor. Successful jobs return a BAMF-hosted MP4 media asset; failed, cancelled, or moderated requests are not reported as successful.

post/generate/talking-videos

Request

  • Base URL: https://api.bamf.ai/v1
  • URL: https://api.bamf.ai/v1/generate/talking-videos
  • Auth: HTTP bearer

Request body

engine'auto' | 'lipsync' | 'veed_lipsync' | 'fabric' | 'omnihuman' | 'kling' | 'speak'

Talking-video engine. Product-launch founder shots use speak, which is Higgsfield-only and does not silently fall back to another video provider.

creator_idstring uuid required
scriptstring required

Words the creator speaks. At most 40 words (about 15 seconds of audio).

image_urlstring uri

Optional HTTP(S) image of the speaker.

quality'mid' | 'high'
promptstring

Optional direction for expression and motion.

metadataobject

Response

Talking video generation job queued

Changes

No recorded changes to this operation across all 1 revision of this API.