Documentation

Everything you need to ship a video.

Quickstart

ViStudioAI turns a structured script into a recorded, captioned, exported video — one scene at a time.

  • Choose your output resolution: 16:9 (landscape) or 9:16 (vertical)
  • Paste your script into the Studio
  • Record each scene, approve or retake
  • Configure CTA overlays, intro clip, and end screen once
  • Merge & export

Open the Studio →

Script format

Each scene is a block separated by a line of dashes. Fields are matched by label, in any order:

Scene 1
Scene Type: video_bg_audio
Visual: Cinematic gold bars slow motion in the dark.
Voiceover: What if I told you this move wasn't about gold at all?
----------------------------------------------------

Recognized Scene Type values: webcam, screen, screen_webcam, screen_webcam_circle, video_bg_audio, notepad, notepad_webcam, custom_video, custom_video_webcam.

Scene types

webcam

Just your camera, full frame.

screen

Screen or window share, full frame, with your own audio plus voiceover mixed in.

screen_webcam

Screen and webcam together — split top/bottom on 9:16, full-screen with a circle overlay on 16:9.

screen_webcam_circle

Always the circle-overlay style (screen full-bleed, webcam as a bubble) regardless of aspect ratio. Trigger it by including "circle" or "overlay" in your Scene Type text.

video_bg_audio

Webcam plus a background image — split left/right on 16:9, top/bottom on 9:16.

notepad

A writable "paper" surface that fills the frame — draw or write live with your mouse while you talk, in any pen color. No webcam.

notepad_webcam

Same writable notepad, plus your webcam — a circle bubble on 16:9, top/bottom split on 9:16.

custom_video

Import a video file and voice over it — no webcam. Recording starts the imported video from the beginning and stops automatically the moment it ends, so the take always matches the clip's length.

custom_video_webcam

Same imported video, plus your webcam as a reaction bubble — the "react to this clip" format. Choose whether the clip's own audio, your mic, or both are included.

16:9 vs 9:16

Pick your resolution once, before recording starts — it's applied to every scene's layout, so switch projects (not mid-recording) if you need both a landscape and a vertical cut.

  • 16:9 (1280×720) — YouTube, LinkedIn, web embeds
  • 9:16 (720×1280) — Shorts, Reels, TikTok

CTA overlays

Configure Subscribe, Like, or social-follow badges from the merge screen — pick the scene, whether it appears at the start or end, and (for socials) your handle. Each badge slides in, plays a notification sound, flips to its confirmed state with a click/shutter sound, then slides out with a whoosh.

Exporting

Merging happens locally using an in-browser ffmpeg build — your scenes, intro clip, and end screen are normalized and stitched together, then downloaded directly. Nothing is uploaded to a render server.

API (Team plan)

Team plan accounts can generate API keys to submit scripts and retrieve renders programmatically. Reach out to enable API access for your account.

curl https://api.vistudio.ai/v1/render \
  -H "Authorization: Bearer vsai_live_xxxxxxxx" \
  -F "script=@my-script.txt" \
  -F "aspect=16:9"

FAQ

Can I use my phone as a webcam?
Yes — apps like DroidCam or Iriun Webcam register as a normal camera device, which the Studio can then select like any other webcam.
Does captioning use speech recognition?
No. Captions are timed against your script's Voiceover text and your recording's actual pacing — so accuracy is guaranteed rather than inferred.
Where are my recordings stored?
Locally in your browser until you export — nothing leaves your device during recording or editing.