Skip to main content

Vision Mode

Vision is the Bankai mode that can see your screen. Hold the Vision shortcut, ask about whatever is in front of you — a chart, an error message, a spreadsheet, a paragraph in another app — and Bankai takes a screenshot, looks at it, and answers. The answer lands right at your cursor, or is read out loud if you ask for it.

It is the difference between describing what you are looking at and simply asking about it. Instead of "there's a number in the third column of a table about Q3 revenue, what's the growth rate," you just say "what's the growth rate here?" while looking at the table. Bankai sees what you see.

note

Vision Mode runs in the Copera desktop app (Windows and macOS), because it needs to capture your screen — something a web browser cannot do. On macOS it also needs Screen Recording permission (see below).

Turning On Vision Mode​

Vision is off until you enable it:

  1. Open the AI section and go to Voice Settings.
  2. In the Modes section, turn on Enable Vision Mode.

Once enabled, Vision joins your mode rotation (its color is magenta and its icon is an eye) and its shortcut becomes active.

macOS: Screen Recording Permission​

On macOS, capturing the screen requires the system's Screen Recording permission. The first time you enable Vision, Bankai walks you through granting it:

  1. Bankai opens a permission step and asks you to allow Screen Recording in System Settings.
  2. After you grant it, macOS requires Copera to restart for the change to take effect — Bankai shows a Restart Copera button to do this in one click.
  3. Once Copera restarts, Vision is ready to use.
note

Vision needs this permission only on macOS. On Windows, Vision works as soon as you enable the mode — there is no extra system permission to grant.

Asking About Your Screen​

Using Vision is the same hold-speak-release rhythm as the rest of Bankai:

  1. Hold the Vision shortcut. Bankai starts listening immediately and grabs a screenshot of your current screen at the same time.
  2. Ask your question about what is on screen — "summarize this," "what does this error mean," "what's the total in this column?"
  3. Release the key. Bankai reads the screenshot alongside your question and answers.

Because the screenshot is captured the instant you start talking, Vision feels just as fast as the other modes — there is no separate "take a screenshot first" step.

Where the Answer Appears​

By default, the Vision answer is pasted at your cursor, just like Ask mode — ready to drop into a document, chat message, or wherever you are working.

If you would rather hear the answer, just ask for it out loud — say something like "tell me out loud" or "read it to me" as part of your question, and Bankai speaks the answer through your speakers instead of pasting it.

Examples​

What you ask (while looking at…)What Vision does
A revenue chart — "what's the trend here?"Reads the chart and describes the trend
An error message — "what does this mean and how do I fix it?"Explains the error and suggests a fix
A spreadsheet — "what's the total of this column?"Reads the numbers on screen and answers
A long article — "summarize this in three bullets"Summarizes what is visible on screen
A form in another language — "translate this to English"Reads and translates the on-screen text

Vision in Omni Mode​

You do not always have to switch to Vision deliberately. If you use Omni mode (where Bankai figures out your intent automatically) and Vision is enabled, Omni will reach for your screen on its own when your question sounds like a screen question — for example, "what's on my screen right now?" or "look at this spreadsheet and tell me the total."

When that happens, Omni captures a screenshot, answers using it, and the result is saved as a Vision session in your history. If your Omni question is not about the screen, no screenshot is taken — so everyday Omni use stays fast.

tip

Keep Vision enabled and stay in Omni mode for the most natural experience: just talk, and Bankai pulls in your screen only when the question actually calls for it.

Your Screenshots Stay on Your Computer​

Vision screenshots are stored locally on your device, not uploaded to Copera. They follow the same retention rules you set for your Bankai audio, so you control how long they are kept (see Voice Settings).

In your Bankai History, Vision sessions show the screenshot that was captured alongside your question and the answer. Click a screenshot to open it full-size in a viewer.

Retrying a Vision Answer​

If a Vision answer was not quite what you needed, you can retry it from your Bankai History — Bankai re-runs the request using the screenshot it already captured, so you do not have to recreate the moment.

Settings and Configuration​

SettingWhat it controlsDefault
Enable Vision ModeTurns Vision on or off (in the Modes section of Voice Settings).Off
Screen Recording permission (macOS only)System permission Vision needs to capture your screen. Granted through a guided step on macOS.Not granted
Screenshot retentionHow long captured screenshots are kept on your device — follows your audio retention setting.—

Tips and Best Practices​

tip

Put the thing you want to ask about clearly on screen before you hold the shortcut. Vision captures whatever is visible the moment you start speaking.

tip

For small text — fine print, dense tables, code — Vision captures at high detail so it can read the details, but a larger or zoomed-in view still gives the most accurate answer.

tip

Use "tell me out loud" when your hands are busy or your eyes are on something else — Vision will speak the answer instead of pasting it.

Frequently Asked Questions​

Why don't I see Vision Mode?​

Vision is off by default — turn on Enable Vision Mode in the Modes section of Voice Settings. It also requires the Copera desktop app; it is not available in a web browser.

Why does Vision not work on my Mac after I enabled it?​

macOS needs the Screen Recording permission, and newly-granted Screen Recording only takes effect after Copera restarts. Follow the guided permission step and click Restart Copera when prompted.

Are my screenshots sent to Copera or stored in the cloud?​

Your Vision screenshots are stored locally on your own device and follow your Bankai retention settings. They are not kept in your notes or uploaded to Copera storage.

Can Bankai read my screen automatically?​

Only when you ask. Vision captures a screenshot when you hold the Vision shortcut, or when you are in Omni mode and your question clearly asks about the screen. It never watches your screen in the background.

Where does the answer go?​

By default it is pasted at your cursor. Ask for it "out loud" and Bankai speaks the answer instead.