You have downloaded a ZIP file. This page takes you from that file to Samsara running on your computer. There is a picture for every step.
The pictures were taken with v0.30.0-beta.1. The steps are the same for any Samsara ZIP — only the version number in the file name changes.
The file is called Samsara-Windows-v0.30.0-beta.1.zip. It is about
570 MB.
Two ways to get to it:
A ZIP is a box. Samsara will not run until you take it out of the box.
Click the file once to select it, then use whichever of these suits you:
A window opens asking where to put the files. The folder it suggests is fine. Click Extract.
Do not run Samsara straight from the ZIP window. If you
double-click the ZIP you can see Samsara.exe inside it, and opening it
from there looks like it should work. It will not. Samsara.exe needs the
_internal folder next to it — 7,740 more files, 1.4 GB unpacked —
and opening it from inside the ZIP copies only that one file somewhere temporary,
without them. Extract first, then run it from the extracted folder.
When extracting finishes, a folder opens containing Samsara.exe and a
folder called _internal. Leave _internal alone — Samsara needs
it.
Double-click Samsara.exe.
Windows will probably show a blue box saying “Windows protected your PC”. This is expected.
Click More info, then click Run anyway.
Why it appears: Samsara is not code-signed. A code-signing certificate is a paid yearly subscription, and Samsara is free software funded by one person. Windows shows this box for every app that has not bought one, whatever the app does.
The first time Samsara runs it asks you seven short questions. Every answer can be changed later in Settings, so do not worry about getting them right now.
A list of what is coming. Click Next.
Pick the description that fits you best. It sets sensible starting options; it does not lock anything.
Choose your microphone from the list, then speak. The Input Level bar should move. If it does not, pick a different microphone and try again.
Extras you can add now or later. Skipping them all is fine.
Bigger models are more accurate and slower. Balanced is the default and is a good starting point.
The keys you will press to talk. The defaults are Ctrl + Shift to dictate and Ctrl + Alt + D for continuous mode. If a key combination is already used by something else on your computer, change it here.
A summary of what you chose. Click Start Samsara.
The speech model downloads the first time you use it. The first sentence you dictate may take longer than the ones after it.
Samsara runs from the folder you extracted. If you would rather not go looking for that folder every time, make a shortcut:
Do not move Samsara.exe itself out of the folder — it needs the
_internal folder beside it. The shortcut is a pointer; the program stays
where it is.
After setup finishes, Samsara opens its main hub window and the first-run tutorial. It also stays available from the system tray.
Nothing here is a silly question. If a step did not go the way this page says it would, the support page lists the things that most often go wrong and how to get in touch.