Skip to content
NeuroDynamic Account Download free

The LexVoca guide

Every page and setting, explained in three layers. Read as much or as little as you need. Prefer to look around first? Take the click-through tour.

What it does in plain wordsWhy it works this wayTechnical detail for the curious

Getting started

Setup guide

Six short steps the first time you open LexVoca. About a minute, and you can change anything later.

What it does

  • Welcome shows your dictation shortcut.
  • How do you work? Three yes / no / not sure questions.
  • We suggest a profile based on your answers - you can pick another.
  • Speech & spelling: where your speech is turned into text, and British or American spelling.
  • A few choices: a handful of switches for your profile.
  • You're all set, with an optional tour.
  • Afterwards, Home shows a short Get started list until you've tried the basics.

Why

LexVoca can do a lot, but most people need only part of it. Asking about how you work - rather than listing features - gets you a sensible setup without reading a manual.

Technical detail
  • Terminals = yes suggests Technical. Otherwise, wanting to use the keyboard less (yes or not sure), or using ClaraVoca or ITSpecta, suggests Assisted. Otherwise Dictate.
  • Spelling starts from your Windows language setting; the guide asks whether to keep it.
  • "On this PC" runs a check of your hardware and gives an estimate - actual performance may differ.
  • Run the guide again any time from Help.

Dictating and the pill

Press your shortcut in any app, speak, press it again. Your words appear where your cursor is.

What it does

  • Start with Alt Z (the default shortcut), a spare mouse button, the green button on the pill, "Start dictating" on Home, or by saying "voca listen" where hands-free listening is switched on.
  • The pill at the bottom of the screen shows what's happening: listening, working, typed, or a problem.
  • Choose press to start and stop, or hold to talk.
  • To change text you've already written, select it, press Alt X and say what to do, such as "make this friendlier".

Why

Typing is the slow part. Speaking straight into the app you're already using means no copying and pasting between windows, and nothing new to learn.

Technical detail
  • With Type where I started on, LexVoca brings the window you started in back to the front before inserting. Keep the intended text field selected: changing fields can stop the text going in. With it off, text goes wherever you are when it finishes.
  • If recognition can't finish - no internet, a sign-in or service problem - the recording may be kept on your PC as an unfinished item in Settings › Speech. What you can recover depends on what was saved and how long you keep unfinished items.
  • Your shortcuts are shown everywhere as they're set on your PC, so if you change them the guide and pages follow.

Profiles

Dictate, Assisted or Technical decides which features LexVoca uses. Simple or Advanced decides how much each screen shows. Switch either at any time.

What it does

  • Dictate: emails, documents and chat.
  • Assisted: use the keyboard less, with help reading and spelling - hands-free phrases, read aloud, ClaraVoca and ITSpecta.
  • Technical: everything, including commands, Superpaste and per-app rules.
  • Simple shows the main controls; Advanced shows the extra settings for your profile. Switch at the bottom of the sidebar.

Why

A screen full of options you'll never use makes the ones you need harder to find. Profiles hide, they don't remove - nothing you set is lost when you switch.

Technical detail
  • Settings a profile turns off are kept aside and given back if you switch back.
  • Tasks & notes is in every profile.
  • When a card holds Advanced settings you've changed, it shows a custom setting count so you can find them in Simple.

Pages

The sidebar, top to bottom.

Home

Ready to go, what needs your attention, and how your dictation is going.

What it does

  • Ready when you are: the LexVoca mark, your shortcut and a Start dictating button. The mark moves while you speak and settles when your words are typed.
  • A small looping screen shows things you can say: writing, terminal commands, Your Voices, opening programs, trigger phrases and reminders.
  • Chips for voice edit, your writing style and your speech connection, with Test and Change links, plus Tidy up what you say. Technical adds Help with commands.
  • Noise: Very quiet room, Normal, Noisy room or Very noisy. It sets when hands-free dictation stops after a pause - choose a noisier setting if it keeps listening, or a quieter one if it cuts you off. It doesn't remove background noise.
  • Needs your attention: an update, kept recordings, or setup to finish - one line each.
  • Your stats: words dictated, dictations, speaking pace and time saved - for this week or all time.
  • Your highlights, by profile. Dictate: up next, where you dictate, your best day and tidied by AI. Assisted adds words learned and response time. Technical adds commands typed and programs opened.
  • Recent: your last three dictations, each with its app's icon.
  • On your first day, a short Get started list: say your first sentence, teach LexVoca a word, and one step for your profile. Steps tick themselves off; hide it whenever you like.

Why

One calm place that answers "is it working?" at a glance. Problems appear only when there are any, and the tiles show what LexVoca is doing for you without opening a settings page.

Technical detail
  • The connection's web address and model are shown only in the Technical profile.
  • Time saved compares your speaking pace with typing at 40 words a minute.
  • Stats are counts kept on your PC - never what you said. Daily figures are kept for 60 days; all-time totals until you reset them.
  • Clearing all History can reset the stats too (the option is off unless you tick it).
  • LexVoca never shows AI allowance numbers. If AI is running low or paused, Home says so in one calm line. Without AI tidying, LexVoca can still use the recognised text; hosted speech needs its own working connection.
  • Noise is the same setting as How quiet counts as finished in Settings › Typing & pasting.
  • The sidebar shows the app version beside the connection light. The light shows the last speech connection check - it isn't proof that every AI feature or a real dictation works.
  • A tile with nothing to show yet stays hidden rather than showing zeros.

History

Your saved dictations and activity, on this PC, newest first.

What it does

  • Each entry shows the app's icon and name, and the time. Filter: All, Dictation, Commands, Programs opened, Tidied by AI. Search at the top.
  • On any entry: Copy, Type it again, Read aloud, Delete (with Undo for five seconds).
  • Where the AI changed your words, "You said" shows the original.
  • Clear all asks first.
  • Say "voca again" (hands-free) to type your newest line once more.

Why

Text sometimes misses its window - a pop-up steals focus, an app is busy. When the recognised text was saved, you can copy it or type it again from here instead of saying it twice. Seeing the original keeps the AI honest.

Technical detail
  • Stored locally only, in your Windows user profile. The most recent entries are kept (200 by default).
  • Speed details - how long it took, where the text came from - are in each entry's details, not the list.
  • Text recovered from an unfinished recording appears here. It isn't typed for you.

Dictionary

Fix words LexVoca keeps mishearing - and, if you switch learning on, let it pick up fixes.

What it does

  • Learning: learn automatically from AI corrections, and (optional) learn when you fix a word yourself.
  • Words for your line of work: load a pack of terms so they're spelt right.
  • Your words: "heard as" → "replace with" rules. Filter by Learned, Added by you, or Check.
  • Export and import to move everything to another PC.

Why

Speech recognition is good at common words and weak at yours - names, jargon, acronyms. A few fixes make a big difference, and learning can add some for you. Recognition can still get words wrong, so check important text.

Technical detail
  • Auto-learn compares what the recogniser heard with the AI-tidied text, and promotes a word fix only after it has seen it more than once.
  • Learned rules are conservative: single words, real-looking terms, never punctuation.
  • Check flags rules that look wrong: spelling-style flips (practising → practicing for a British user) and a word turned into its plural.
  • LexVoca won't learn spelling-style flips in the first place.
  • "Learn when I fix a word" reads only the text box you just dictated into, read-only, for about 25 seconds, and never in terminals or consoles.
  • Word packs are hints sent to the recogniser; your own list is applied to every transcript.

Tasks & notes

Speak a task, note or reminder. LexVoca files it in Tasks & notes, ready for you to check.

What it does

  • Capture with Alt Shift Z (the default shortcut), the microphone button, or by typing.
  • Tasks are grouped Overdue, Today and Later. Tick to finish, Snooze, or Edit the wording, day, time and repeat.
  • Repeat: never, every day, every week or every month.
  • Notes are colour-tagged cards: read aloud, edit, make into a task, delete.
  • When a task is due, a reminder pops up on screen, as long as LexVoca is running and your PC is awake. Check the day and time LexVoca worked out from your words.

Why

Ideas and to-dos arrive while you're doing something else. Saying them out loud, without switching apps, means they're captured before they're forgotten - and the reminder comes to you.

Technical detail
  • "Remind me to..." makes a reminder, and days and repeats ("tomorrow", "every Sunday") are picked up from your words. Check the time in Tasks & notes and change it in the Edit view if needed.
  • Reminders appear on top of other windows even when LexVoca is minimised or in the tray, without taking your keyboard focus. LexVoca must be running and the PC awake.
  • Ticking a repeating task moves it to its next date rather than finishing it.
  • Each note's colour comes from its topic, so related notes look alike - a small memory aid.
  • Everything is stored on this PC.

Trigger phrases

Say a short phrase, get a whole block of saved text.

What it does

  • "Email signature" types your sign-off. "Standup update" types your template.
  • The text can span several lines.
  • Start from an idea: signature, standup, out of office.

Why

A lot of what we type is the same every time. Two spoken words are faster than any template menu, and work in every app.

Technical detail
  • The phrase has to be the whole dictation, so it won't fire in the middle of a sentence.
  • Command phrases for terminals live on the Commands page; words after a command phrase are added as arguments.

Commands

Technical

In a terminal, say what you want. LexVoca types the command line for your shell, and you check it before pressing Enter.

What it does

  • Three ways to say it: spell it (in Bash, "L S dash L A" → ls -la), say the words (in PowerShell, "get child item" → Get-ChildItem), or just ask (in Bash, "show disk space" → df -hT). The shell decides the command.
  • Try a command for Bash, PowerShell, Command Prompt or VMware ESXi, then save it as a short phrase.
  • Which windows get commands, words LexVoca hears wrongly, and a searchable reference of everything it understands.

Why

A command typed into the wrong place can do damage, so LexVoca is deliberately careful: it sticks to an allowed list of commands that look around or move between folders, plus phrases you saved yourself, and you read the line before it runs. The checks help, but they aren't a safety guarantee.

Technical detail
  • Only recognised command words can start a line, so a sentence spoken into a terminal by mistake is never typed.
  • Anything that deletes or changes files is refused unless you saved it yourself as a phrase - and saved phrases can change things, so check them.
  • PowerShell and Command Prompt are recognised by window title. PuTTY or SSH windows look the same whatever's at the far end, so LexVoca asks once - say 1 Bash, 2 PowerShell, 3 Command Prompt or 4 normal text - and remembers per program or per user@host.
  • If the built-in grammar can't make a command, the AI can suggest one line. It's checked against the same list and marked "AI:".
  • Auto-enter is off by default. When on, it presses Enter only for supported checks and folder moves in other shells. Bash always waits for you to review the line and press Enter.
  • Symbols: "dot", "slash", "dash", "space" for a new argument, "literal" to keep a word, "spell ... end spell" for letters.

Launch

Assisted and Technical

"Open my emails" opens Outlook. "Open downloads" opens the folder. "Launch writer" starts a new document.

What it does

  • Try a name shows what would open and why - it never opens anything.
  • Words that start a program ("launch", "open") and words that bring it forward ("switch to", "go to"), as chips you can add or remove.
  • Programs are listed by type, each with its own icon. Give them your own spoken names, and add apps that aren't in the Start menu.
  • Folders: "open downloads", "open my documents", or folders you name yourself. In a terminal, "go to logs" and other named places are on the Commands page.
  • Ask before opening, if you'd like to confirm first.

Why

Finding a program in the Start menu takes several clicks. Saying its name - or your name for it - takes a second.

Technical detail
  • LexVoca can only open Start menu shortcuts or apps you add, so it can never run something that isn't installed.
  • If several programs match, the pill shows numbered choices - say the number or "cancel".
  • If the program is already open, LexVoca brings its window forward instead of opening another.
  • Rescan after installing something new.
  • Folders open in File Explorer, and only folders that exist on this PC.

Apps & websites

Technical

For windows where your words don't arrive properly - consoles in web pages, remote sessions that ignore Ctrl + V.

What it does

  • Check a window: click into the problem window and LexVoca shows how it puts text there, and why.
  • Pick a method and remember it for that website or that app:
    • Normal paste - Ctrl + V, right for most apps.
    • Terminal paste - Shift + Insert, for web terminals and SSH.
    • VM console keys - for VMware, noVNC and consoles that ignore paste.
    • Type each key - when nothing else works.
  • Lists of the apps and websites you've taught it.

Why

Windows can't always tell LexVoca what kind of window it's typing into - a remote console looks like an ordinary browser tab. Teaching it once fixes that window for good.

Technical detail
  • Order of decision: a website rule (closest matching address wins), then an app you've taught it, then automatic.
  • LexVoca reads only the address bar of Edge, Chrome, Firefox, Brave, Opera and Vivaldi - never the page or your history.
  • Anything after ? or # in an address is ignored. If a browser hides http/https, the rule matches either.
  • If two rules match equally, or the address can't be read, the app setting is used.
  • Rules apply in Automatic mode, to dictation and to the clipboard Superpaste.

Help and reading options

Short answers inside the app, read your way.

What it does

  • Topic cards and a search box: "What do you want to do?"
  • Each topic opens on its own page: the gist, numbered steps, tips, and links straight to the right settings.
  • Reading options: font, size, line and letter spacing, line length, background tint, text colour and bold word starts.
  • Read aloud with ClaraVoca. Run the setup guide again.
  • Feedback, beside Help: tell us what went wrong or share an idea. You can add your last few dictations and see exactly what is sent. Logs aren't sent; use Export diagnostics in Backup & more instead.

Why

Everyone reads differently. Letting you shape the text - rather than giving one "accessible" version - means help that's comfortable for you.

Technical detail
  • Reading options are saved with your settings and apply to every help topic.
  • Read aloud needs ClaraVoca, our reading companion app, and the Assisted or Technical profile.

Settings

Seven tabs, one topic each. "Find a setting" searches them all.

Dictation

How LexVoca treats what you say, and which language it writes in.

What it does

  • Automatic (recommended): commands in terminals, ordinary text everywhere else. Or Always text, or Always commands.
  • Work in administrator windows. Ask the AI when a command isn't recognised. Auto-enter (use with care).
  • Spelling: British or American.
  • Write my dictation in another language - speak one, type another.
  • Translate speech to English (own server only).

Why

The same words mean different things in an email and a terminal. Automatic mode looks at the window you're in, so you don't have to switch modes by hand.

Technical detail
  • Typing into windows opened "as administrator" needs LexVoca to run with administrator rights too; the switch sets that up.
  • Spelling starts from your Windows language. The AI tidy-up and translation follow it.
  • Translation keeps names, numbers, code, e-mail and web addresses exactly as you said them. Commands in a terminal are handled before any translation.
  • Some options only appear in the Technical profile; the tab says so.

Shortcuts

The keys and mouse buttons that start LexVoca, and whether they're working.

What it does

  • NeuroDynamic shortcuts: hold Left Alt and press one key. LexVoca uses Alt Z to dictate and Alt X to change the selected text by voice. Add Shift for the bigger version: Alt Shift Z speaks a task or note, Alt Shift X is Super Paste.
  • Alt Q opens the family menu: LexVoca, ClaraVoca and ITSpecta actions that have no key of their own.
  • Each shortcut shows a status: Working, In use, or Not checked. Change any of them. The page also lists the keys your other NeuroDynamic apps use, and warns before you pick one.
  • In the family menu, LexVoca's items are H open LexVoca, P pause or resume listening, and I dictation history. Press the letter or click.
  • Inside LexVoca, as in every NeuroDynamic app: Ctrl K search, Ctrl , Settings, F1 help for the page, Esc to close, and Alt 1-9 for the pages in the sidebar.
  • Press to start and stop, or hold to talk - separately for each.
  • A spare mouse button to dictate, and whether to keep that button for LexVoca only.

Why

One set of keys for all three apps, in one corner of the keyboard, means less to remember. Another program may still use a key, and then nothing happens when you press it. Showing the status means you find out here, not mid-sentence.

Technical detail
  • "Check the shortcuts" tests whether each combination can be registered.
  • Use the left Alt key: on many keyboards right Alt is AltGr and types characters instead.
  • NVIDIA's overlay also uses Alt+Z. Change it in the NVIDIA app if LexVoca reports Alt+Z "In use".
  • When you update, LexVoca offers the new shortcuts once and keeps any keys you chose yourself unless you accept. See every NeuroDynamic shortcut.
  • Mouse buttons are useful when a virtual machine has grabbed the keyboard - the mouse isn't part of that grab.
  • Keeping the button for LexVoca only needs a system-wide mouse hook, so it's off until you turn it on.

Typing & pasting

How your words get into the window, and the finishing touches.

What it does

  • Automatic (Superpaste), Always paste, or Always type.
  • Superpaste the clipboard into a VM with Alt Shift X (the default shortcut) or a mouse button.
  • Finishing touches: type where I started, put my clipboard back, add a space after, and end a sentence only when you say "full stop" (experimental).
  • Hands-free voice: "voca listen", "voca again" and "voca paste" by default, or phrases you choose. How quiet counts as finished is the same setting as Noise on Home.
  • Work inside VirtualBox (Technical): an optional workaround for VirtualBox consoles only.

Why

Pasting is fast and works almost everywhere. Where it doesn't - consoles, VMs - LexVoca types instead, so you don't have to think about it.

Technical detail
  • Automatic types into remote, VM and console windows and pastes elsewhere. Per-app and per-website overrides live on Apps & websites.
  • Your clipboard is put back after pasting, unless you turn that off.
  • Hands-free phrases are heard on this PC only, using the Windows default microphone. Whether LexVoca listens depends on your profile, your settings and the window in front.
  • Typing speed in consoles can be slowed for laggy remote sessions.

Speech

Where your speech is turned into text.

What it does

  • NeuroDynamic hosted: fast and accurate, with your NeuroDynamic account.
  • Provider or own server: a speech provider such as OpenAI or Groq with your own key, or your own server. Your provider's limits apply. Quick setup, saved addresses, Test.
  • On this PC: speech is recognised on your computer. Download a speech model, check this PC, measure its speed. Terminal commands aren't available with this choice.
  • This choice is for speech only. AI features use your NeuroDynamic account whichever you pick.
  • Three different checks: Test connection checks access and model information; Start speech test records a real test; Check AI access, in AI, checks your account's AI.
  • Unfinished recordings: shown only when there's something to review. Recover, Process again or Delete each one, and choose how long they're kept. Recovered text goes to History; it isn't typed for you.

Why

People need different things: the best accuracy, speech that stays on their PC, or their own server. Each choice keeps its own settings, so switching back is instant.

Technical detail
  • On this PC runs a speech model locally. LexVoca checks for a usable graphics card first and can fall back to the processor. "Check this PC" estimates whether your hardware can run it - speed and accuracy depend on the model and your hardware.
  • Model sizes come from the live list in this tab; models change over time.
  • Own server: any OpenAI-compatible speech endpoint. Spare addresses for the same server are tried only if you enter them.
  • Keys are kept in Windows' protected storage, not in plain settings files.
  • Want to run your own speech server? Our step-by-step guides show you how.

AI

Tidying, writing styles and voice edit, through your NeuroDynamic account.

What it does

  • Your NeuroDynamic account: AI features run through it. ClaraVoca and ITSpecta use the same sign-in, and signing out signs you out of all of them.
  • One switch for AI, plus switches for tidying and help with commands. Your Voices, voice edit and translation are set where you use them. Try it shows the tidied result first.
  • Write my dictation in another language Pro is part of the Pro plan.
  • Signed in: Check AI access, Manage account and Sign out. Signing in opens your browser; while it waits you can cancel or continue there. Signing in to the website and connecting the app are separate, so LexVoca may offer Reconnect.
  • Calm states, never numbers, on Home and the pill: AI running low, with the reset date; AI paused until a date; Confirm your email; Service paused, with a link to the status page. Signed out, it says Sign in to use AI features.
  • Your Voices: give an app a writing style - professional for Outlook, friendly for Teams, clear and formal for Word - or keep Just as I say, which means no style rewrite. Learning examples aren't saved by LexVoca; a preview is saved with a style when you press Save.
  • Edit the cleanup instructions, or reset them.

Why

Raw speech-to-text has no punctuation and plenty of "um"s. A light tidy makes it ready to send - but your meaning must never change, so the AI is held to strict rules. One account means nothing to set up: NeuroDynamic picks the right model for each job.

Technical detail
  • If AI tidying is unavailable, LexVoca can keep the recognised text instead. Hosted speech still needs sign-in, and voice edit and other AI-only actions need AI access.
  • Your speech stays where you chose - on this PC, your own server or NeuroDynamic. Only the text goes to NeuroDynamic AI, and only for features you have on.
  • The AI is told to make minimal corrections only. Its answer is checked, and if it seems to have changed names, numbers or meaning, the original text is kept. The checks catch some changes, not every one, so review important text.
  • Very short dictations skip the AI - nothing to tidy, and faster.
  • Your spoken words are always treated as text to tidy, never as instructions to the AI.
  • Spelling follows your British or American setting.
  • Styles apply to eligible dictation and skip protected boxes such as passwords. History keeps what you said. Voice edit (select text, say what to change) uses the app's style too.

Appearance

Theme, the pill, the tray icon, your microphone and sounds.

What it does

  • Theme: Dark (default), Light (warm cream with orange accents), or Match Windows.
  • Keep the pill on screen, and choose its quick actions: seven chips, each with its own symbol and colour. Pick which ones show, and change any colour.
  • Always show the tray icon; the plain Windows tray menu for screen readers.
  • Microphone choice, sounds, and play a sample.

Why

A comfortable look matters if you use LexVoca all day. The quick buttons put what you do most one click away.

Technical detail
  • Both themes meet contrast guidelines: text at least 4.5 : 1.
  • Animations switch off when Windows' "reduce motion" setting is on.

Backup & more

Back up your settings and words, ClaraVoca, and updates.

What it does

  • Choose what to include, then back up to a file or to your NeuroDynamic account - and restore.
  • ClaraVoca status and a link to open it.
  • Updates: install and restart, or view the update logs.
  • Export diagnostics (Advanced): saves logs and the app version to a file you can share with support. Dictated text, recordings, History, settings and passwords are left out, and nothing is uploaded.

Why

Your dictionary and phrases take time to build. A backup means a new PC - or a reinstall - starts where you left off.

Technical detail
  • Keys, passwords and Auto-enter are left out of backups. Phrases and History you choose to include may contain personal text.
  • Words, vocabulary, trigger phrases and app tones can also be moved on their own from the Dictionary.

Good to know

Privacy and your data

What stays on your PC, what's sent, and what LexVoca never looks at.

Stays on your PC

  • History, tasks and notes, your dictionary, trigger phrases and settings.
  • Unfinished recordings, until they're recovered, deleted or expire.

Sent, only for what you use

  • Your recording, to the speech service you chose - not at all with On this PC.
  • The recognised text, to NeuroDynamic AI, for AI features you have on: tidying, Your Voices, voice edit and translation.

Never

  • Page contents or browsing history - website rules read only the address bar.
  • What you type - LexVoca watches only for its own shortcuts and never records your typing.
  • Terminals or consoles, for "learn when I fix a word".

Troubleshooting

Quick fixes for the most common problems.

Nothing happens when I press the shortcut
  • Open Settings › Shortcuts. If it says "In use", another program has that combination - choose a different one.
  • Use the left Alt key. On many keyboards right Alt is AltGr and types characters instead.
  • If Alt + Z opens the NVIDIA overlay, change the overlay's shortcut in the NVIDIA app.
  • Inside a virtual machine, the VM may grab the keyboard. Try a mouse button, the tray menu or Paste in 3 seconds. For VirtualBox, Work inside VirtualBox can help.
LexVoca can't hear me
  • Check the microphone isn't muted and the right one is chosen in Settings › Appearance.
  • "No speech detected" can also mean the recording was too quiet or too short. Try a full sentence, close to the microphone.
Hands-free dictation stops too soon, or keeps listening
  • Change Noise on Home: a noisier setting if it keeps listening, a quieter one if it cuts you off. It doesn't remove background noise.
My words didn't appear
  • If the text was saved, it's in History - Copy it, or Type it again. If recognition didn't finish, look for an unfinished recording in Settings › Speech.
  • Turn on Type where I started, and keep the text field selected while you speak.
  • If it keeps happening in one app or web page, use Check a window on Apps & websites and pick another method.
  • For windows opened as administrator, turn on "Work in administrator windows" in Settings › Dictation.
A word is always wrong
  • Add it in the Dictionary: what LexVoca hears, and what it should type.
  • Load the word pack for your line of work.
The spelling is American (or British) and I want the other
  • Settings › Dictation › Spelling. Then use the Dictionary's Check filter to remove any old fixes that flip it back.
It says a recording was kept
  • Speech recognition didn't finish. The message says whether it was the connection, sign-in, the service or the speech model.
  • Review it in Settings › Speech under Unfinished recordings: Recover puts the text in History, where you can copy it or type it again.
A reminder didn't pop up
  • LexVoca must be running - it can be minimised or in the tray, but not closed - and the PC must be awake.
  • Check the task has a day and time: open it with Edit.
LexVoca typed a command when I wanted words
  • Say "normal text" in that terminal, or set it on the Commands page under Which windows get commands. Or choose Always text in Settings › Dictation.