Welcome
Workbench is made to be found by feel. Here’s the lay of the land and the few things worth trying first.
Set up
Free, with no account. The speech model downloads once, then works offline.
For Macs with M1 or later, on macOS 14 or newer. The iPhone, iPad and Chrome editions are paused while the Mac app comes first.
Finding your way
The same tools and saved work, wherever you start them.
Every tool has a keyboard shortcut. See them all, change them or practise them in Settings → Keyboard.
Voice to Text
Click the field you want to use, start Voice to Text and speak. In supported fields, Workbench can add words as you speak or paste the completed transcript. If the field or permission prevents automatic paste, use Copy and paste the words yourself. Every transcript stays in History. Moving or editing the field stops live insertion, and the complete words are copied so you can replace what was written.
Meetings
Meetings is a separate, privately distributed app. People with repository access can download Meetings; it is not included in the Workbench download. Once installed, use Import from Workbench to copy previous recordings and transcripts from the edition you used. Meetings asks for its own permissions when you start recording.
Your originals stay in place. Without Meetings, earlier transcripts remain readable in Workbench’s History, and Settings → General → Show saved recordings opens the original files.
Snap
Choose Region, Window or Screen. Copy or export the result, or crop and annotate it first. Save keeps it in History without changing the original.
For another app, use Copy image and paste, or save a PNG with Export image. You can also try dragging a saved thumbnail or the editor’s Copy image control. Drag delivery has not been verified for this release; use Copy or Export if the image does not arrive.
Enable it in Settings → General → Experimental features. Capture screens with narration, review the session, then hand it to an assistant to build a deck. It stays inactive while disabled, and existing session folders are preserved.
Annotate
Circle, arrow or sketch over whatever is on screen, then press Escape and carry on. Whiteboard and Timer live here too.
Present · Persona
Your photo, a live camera bubble or a role card, floating over any app. Its ring can move as you talk.
History and Hand off
Voice transcripts, Snaps and assistant results in one searchable place. Select a few and hand them to an assistant; the result comes back here.
Review transcript shows the full current text and original wording without replacing your Voice to Text draft. Search can find words inside an assistant’s saved answer. Open Review result to copy its exact words. Read text-to-speech is retired; an old Read draft is preserved as a text file in Library. Existing exported audio stays where you saved it.
Library → Resources keeps reusable prompts, links and file references. Search or show favourites, then copy the complete prompt or open the selected link or file. If a file moved, Locate file lets you review a replacement before saving it. Full paths and Change file are under File details; original files stay in their own folders. Text beyond the prompt limit is refused with guidance to save it as a file, never shortened silently.
Team packs
Library → Packs keeps installed content ready offline. Choose Add a pack to enter a repository and connect the GitHub account your team invited. Choose the tool for a skill, or save a resource into your Library. Your own copies never change when the pack updates.
Present
Pick a scene, plug in your iPhone (unlocked, trusting this Mac) and choose the Present window in your call app’s sharing controls. Check the participant’s view before the demo. The phone shows as video; you tap and talk on the phone itself.
Start from a starter or your own photo, add your logo, and drag the phone and backdrop into place. Scene options holds the rest: the device frame, logo, crop, device shape and hand. Every scene is a still picture you can also export.
Help