Getting started

Install Startwell Capture on macOS, Windows or Linux, grant permissions and make your first capture.

Install

Startwell Capture runs on macOS, Windows and Linux. Windows and Linux are new: the app is the same, and where something works differently there, these pages say so. Each system installs with one line in its terminal. The install page picks yours and has direct downloads too.

macOS

You need a Mac with Apple silicon (M1 or later) on macOS 15 Sequoia or later. Open Terminal (in Applications, Utilities) and paste this line:

Paste into Terminal
curl -fsSL https://capture.startwellventures.com/install.sh | bash

It downloads the latest version, checks it, and puts Startwell Capture in Applications. Before macOS 26 it also fetches the speech model. Run it again any time to update. Read the script first.

It downloads the latest release, checks it against the published SHA-512 and our signing certificate, puts Startwell Capture in Applications and opens it. There is no “Open Anyway” step, and the permissions you grant are kept when the app updates. Run the same line again any time to update by hand.

Prefer not to use Terminal? Download the .dmg instead. Because the app is not notarised by Apple, that route needs one extra click in System Settings the first time; the install page walks through it.

Windows

You need Windows 10 or 11, 64-bit. Open PowerShell (search for it in the Start menu) and paste this line:

Paste into PowerShell
irm https://capture.startwellventures.com/install.ps1 | iex

It downloads the latest version, checks it, installs Startwell Capture for you (no administrator needed) and fetches the speech model. Run it again any time to update. Read the script first.

It installs the app for your account only, with no administrator rights. The Windows build is not code-signed yet, so a downloaded installer gets a SmartScreen warning; the command avoids it. Windows asks for no permissions up front, but the microphone has to be switched on for apps in Settings, Privacy & security, Microphone.

Linux

You need a 64-bit Intel or AMD computer with glibc 2.39 or later (Ubuntu 24.04, Debian 13, Fedora 40 or newer), on X11 or Wayland. Open a terminal and paste this line:

Paste into a terminal
curl -fsSL https://capture.startwellventures.com/install.sh | bash

It downloads the latest version, checks it, installs the .deb on Ubuntu and Debian or the AppImage anywhere else, and fetches the speech model. Needs glibc 2.39 or later: Ubuntu 24.04, Debian 13, Fedora 40 or newer. Run it again any time to update. Read the script first.

On Ubuntu, Debian and other apt systems it installs the .deb, so it asks for your password. Everywhere else it installs the AppImage, which needs FUSE 2. On an older system it stops before installing anything. On Wayland your desktop asks which screen to share until you choose one in Settings, Capture, and some context is not available; see Troubleshooting.

Transcription

Speech is turned into text on your computer. On macOS 26 and later that is Apple’s own transcriber. On Windows, Linux and earlier macOS it is Whisper, and the install command downloads the speech model that suits your computer. Either way the audio never leaves it.

The app lives in your menu bar on macOS and in the system tray on Windows and Linux. Its main window opens from there, or with ⌘⇧K (Win+Shift+K on Windows, Super+Shift+K on Linux).

The welcome

The first time it opens, a short welcome sets out what the app does, checks the permissions it needs, and asks you to try one capture. You can skip it and bring it back later from the shortcuts sheet.

The welcome: Capture now, hand it to your AI later. Four rows: capture from any app, Focus is where captures go, packs hand it to an AI, walkthroughs for testers.

Your first five minutes

  1. Press ⌘⇧TWin+Alt+Shift+TSuper+Shift+T, type a note and press Return. A card at the top right confirms it.
  2. Press ⌘⇧SWin+Alt+Shift+SSuper+Alt+Shift+S and drag over part of the screen. macOS asks for Screen Recording the first time. On Windows it is Win+Alt+Shift+S, because Win+Shift+S is Windows’ own Snipping Tool.
  3. Press ⌘⇧VWin+Alt+Shift+VSuper+Shift+V, say a sentence, press ⌘⇧VWin+Alt+Shift+VSuper+Shift+V again. The transcript, with what was on screen, lands on your clipboard when it is ready. macOS asks for the microphone the first time; on Windows, check the microphone switch above.
  4. Make a topic for something you are working on, then set it as your Focus. New captures now go there. See Focus and groups.
  5. Open the main window. The Inbox lists today’s captures, grouped by when you took them.
The main window with the Inbox open.
The Inbox: one line per capture, the selected one on the right.

Around the main window

  • Capture from any app, bottom left: every capture with its keys. On macOS hold ⌘⇧ and press the letter, from anywhere. On Windows most are Win+Shift and the letter, but S, V, T and M are Win+Alt+Shift because Windows keeps those, and a link is Win+Shift+O. On Linux they are Super+Shift, with Super+Alt+Shift+S for a screenshot; on Wayland they work only while the app is in front. See Shortcuts.
  • Hints: a one-line tip appears the first time it is useful, and stays gone once closed.
  • Press ?, or ⌘/ (Ctrl+/ on Windows and Linux), for the shortcuts sheet.
  • Focus at the top left: the topic new captures go to.
  • Inbox: everything, newest first, in groups.
  • Unsorted: captures without a topic, each with a suggestion. See Sorting.
  • Search and the ⌘⇧K palette: every word in notes, transcripts, screenshot text and links.
  • Walkthroughs: record, review and share. See Walkthroughs.
  • Issues: the bugs, requests and ideas raised in your recordings. See Issues and hand-off.
  • Hand off: one prompt for your agent from your recent recordings, and your saved prompts.
  • Graph: how your topics connect. It shows only once you switch on Show the Graph in Settings, General.
  • Topics: each has its captures, memory, outputs, rules and agent access.

AI, if you want it

Everything above works without AI. Titles come from free rules, or from a model on your computer: Apple’s on-device model on macOS, or the optional Smart titles and sorting download on Windows and Linux. Summaries, digests and topic memory use a model of your choice through OpenRouter, with your own key:

  1. Create a key at openrouter.ai/keys.
  2. Paste it into Settings, AI, and pick a model.

The key is stored encrypted on your computer, with the encryption key in your macOS Keychain, in Windows’ data protection for your account (DPAPI), or in your Linux keyring. Requests go straight from your computer to OpenRouter. You can set a monthly spending limit. AI features stay hidden until a key is added.

Where your data lives

  • The database, media, settings and speech models: ~/Library/Application Support/Startwell Capture on macOS, %APPDATA%\Startwell Capture on Windows, ~/.config/Startwell Capture on Linux.
  • The context packs, one folder per topic: ~/Startwell Capture on macOS and Linux, %USERPROFILE%\Startwell Capture on Windows. You can move it in Settings, Library.

We hold no copy. Deleting those folders removes everything.