Video Studio

Plan a video series with the assistant, edit and approve its scripts, narrate and render it with screen recordings, music and captions, then publish versions and download them.

Copy page for AIUpdated November 21, 2026

A short narrated video often teaches a task better than a page of text, but few teams have time to film and edit one. At Harbor Point Logistics, the operations manager asks for a two-minute video for new dispatchers on logging a delivery, approves the script, and has a captioned video the same afternoon.

Add it from the Marketplace

What it does

Video Studio makes narrated videos from a chat: training videos, status updates, progress reports, presentations and short marketing pieces. You tell the assistant who the video is for and what it should cover, and it writes the script scene by scene beside the chat.

Once a person approves the words, the studio narrates them, places your screen recordings, adds music and captions, and joins the scenes into one video.

Set it up

  1. Add Video Studio from the Marketplace. See Marketplace: add or remove a module.
  2. Choose the speech provider. ElevenLabs voices and music need your workspace's own ElevenLabs key under System Settings → AI Providers. OpenAI and xAI voices use your workspace's AI providers. See AI providers.
  3. Open Video Studio settings from the ⋮ on the module's sidebar row. See Module settings and versions.
  4. Under Narrator voice, pick the default voice. Record my voice makes a voice that sounds like you, after you read a consent statement and a short passage aloud.
  5. Under Brands, add your logo, colors and font.
  6. Under Pronunciation, add your abbreviations with Written and Said as.
  7. Decide who approves scripts and who narrates and renders. See Users and roles: roles and permissions.

Plan and write a video

  1. Click Video Studio in the sidebar. It opens on Home, with every series as a card and the assistant's chat beside it.
  2. Tell the assistant who the video is for and what it should cover. Paste notes if you have them.
  3. Agree its plan: how many videos and roughly how long. The scenes appear on the timeline as it writes.
  4. Click a scene. The right panel has Script (what the narrator says), Visual (layout, items and screens) and Timing.
  5. Edit the words yourself, or ask the assistant to "make this tighter". It rewrites only the open scene.
  6. In the ⋯ menu, choose Check script and fix any words the narrator may misread.

To show your application's screens, ask the assistant to record them. It shows its plan first, and recordings never press Save, Submit or Send. New series on Home starts another series.

Approve, narrate and render

  1. Choose Approve script. This is a person's sign-off on the words.
  2. On the Script tab, choose Narrate. If the series has no narrator yet, pick one and narration starts.
  3. Choose Render video. A ribbon across the top shows progress. You can keep working, or leave and come back.
  4. When it is done, choose Watch video.

Any change to a scene's words, layout or items sends the script back to draft. Only scenes whose words changed are voiced again. If a long narration stops with "session expired", choose Narrate again and it carries on.

Music, versions and whole series

Music. Choose Sound at the top, then Make it to make a track from a description, Upload a track for an MP3, WAV or M4A you have the right to use, or Your library. Set Music volume, and keep Quieter while someone talks on. Volume changes never send a script back to draft.

Publish and download.

  1. Watch the finished video.
  2. Choose Publish this render, under the video or in the ⋯ menu.
  3. It becomes the next version: 1, 2, 3. A published version never changes.
  4. The Version button at the top downloads the latest: the 1080p video, a 720p web copy, captions and the poster.

Whole series under the timeline shows every video end to end; Render joined video makes them one video.

Who can do what

PermissionWhat it lets someone do
View videosSee series, scripts, renders and published versions; watch and download
Write and edit scriptsCreate series and videos, edit scenes, upload recordings and music, set volumes
Approve scriptsApprove a script before narration, and publish a render
Narrate and render videosNarrate, render, and make music, pictures and clips
Manage voice consent recordsRemove a studio voice's stored consent recording. Nobody has it by default

Ask its agent

The Video Studio assistant writes and changes scripts, checks them, suggests music and reports on renders. Try:

  • "Make a three-video onboarding series for new staff."
  • "Write a two-minute status update from this list of what shipped."
  • "Check the scripts for words the narrator will misread."

It asks before narrating, rendering or making music or pictures, and never approves or publishes unless you ask. See Agent chat.

Good to know

  • Renders take time. A seven-minute video takes about half an hour, and editing meanwhile is safe: the render uses the script as it was when it started.
  • Use only music you have the right to use, especially in videos you publish.

More help

Open Help in the module for step-by-step guides and common questions. The Knowledge Base module can send an article or a whole section here as a draft series. To show Support a problem instead, see Record your screen and send it to Support. See also Module help and Show me.

Can't find what you need? Contact support.