Skip to main content
GET
Get visual captions (4 modes)
Product: Video Datalake Host: https://api.memories.ai/datalake/v1 Auth: Authorization: sk-mai-...
Read visual captions (generated by OmniCaptioner). Four modes by precedence as=url > since > start/end > no params: whole text, a time window, an incremental cursor (live polling), or a WebVTT file link.
Pricing: $0.001 / call · See Pricing for the full model.

Response Fields

string
Whole-text mode: the full caption string.
array
Windowed / incremental: [{start,end,text}].
string
Incremental mode cursor.
boolean
true while still indexing (more segments coming).
string
as=url mode: signed WebVTT link.

Notes

Empty content while still processing409 video_not_ready.

Authorizations

Authorization
string
header
required

Send Authorization: sk-mai-...

Path Parameters

id
string
required

Video ID.

Query Parameters

start
number

Window start seconds.

end
number

Window end seconds.

since
string

Incremental cursor (empty for first page).

as
string

url to get a .vtt signed link.

Response

200 - application/json

Success