Video generation is slow and expensive, and the two moments you need control are before it starts and while it is running. Estimating first turns a guess into a decision; being able to cancel means a run started on a typo does not have to finish. The grammar helper is a small touch that reflects how these fail in practice — the script is read aloud, so an awkward sentence becomes an awkward delivery you cannot fix without regenerating.
VlogMe generates video from a script using voices and portraits. The endpoint covers the full run — including estimating what it will cost and stopping it if you change your mind.
- Preparation — script_grammar_help, list_voices, list_portraits
- Cost — get_balance and estimate_credits before you commit
- Generation — generate_video, get_video, cancel_video, list_my_videos
- Support — reporting a bug and tracking its status, from inside the connection
OAuth or a bearer token, exclusively one or the other. The token is displayed once, and their documentation is explicit about handling it as a secret.
Paste a URL, then authorize — add the endpoint to your client, then approve the OAuth consent screen