Skip to main content
POST
Create take
Create a take on a clip and start audio. Send avatar_id and exactly one of text or phonemes. Optional name defaults to Take N. If text is sent, it is converted to phonemes, then synthesized. Pass phonemes for pronunciation control. Do not send both. Or send from_take_id (and nothing else except optional name) to copy audio, phonemes, and video instructions from another take on the same clip. That copy does not charge extra credits. The source take must already have audio. Credits are charged when audio starts from text or phonemes. If there are not enough credits, the request fails and no take is created. Poll Get take until status is ready_to_render and audio_url is set. Full flow is in Workflow.

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

clip_id
string<uuid>
required

Clip ID.

Body

application/json
avatar_id
string<uuid>

Required unless from_take_id is set.

text
string

English line to speak. Converted to phonemes, then audio.

phonemes
string

Pronunciation string. Do not send with text.

from_take_id
string<uuid>

Copy audio, phonemes, and video instructions from another take on this clip. Do not send with text, phonemes, or avatar_id.

name
string

Take name. Defaults to Take N.

Response

Take created; audio is starting

id
string<uuid>
required
clip_id
string<uuid>
required
name
string
required
status
enum<string>
required

Current landmark. Poll until ready_to_render for audio, complete for video.

Available options:
starting_tts,
generating_audio,
generating_motion_instructions,
ready_to_render,
starting_video,
preparing_video,
generating_face_video,
generating_avatar_gestures,
compositing,
applying_labels,
watermarking,
signing,
complete,
failed
avatar_id
string<uuid>
required
error
string | null
required
audio_url
string | null
required

Hour-long download URL once audio exists.

video_url
string | null
required

Hour-long download URL once video exists.

created_at
string<date-time>
required
updated_at
string<date-time>
required
text
string
phonemes
string
chunk_prompts
object[]
show_ai_label
boolean
estimated_seconds
number

Approximate seconds left in the current job.

credits_charged
any

Credits charged for this request, when a job was billed.

credits_used
any
credits_remaining
any