Preview — payments disabledYou can create an account and use the site, but checkout is switched off and no subscription can be started here.
Skip to main content
Xuula

A voice productivity workspace for Windows

Speak anywhere. Keep every meeting. Send writing, not a transcript.

Xuula types clean mixed-language text into the apps you already use, records meetings with a live speaker-separated transcript, and turns finished conversations into summaries and to-dos — kept together in Projects.

2,000 words free every week. No credit card.

Windows 10 and 11 · 64-bit (x64)

Давай перенесём the design review на четверг — я отправлю the updated roadmap сегодня вечером.
F8Hold to talk. Release to insert.
An illustration of a message window with dictated text being inserted into it.
Listening

“Up to 4x” compares typical speaking speed with typical typing speed. Your own result will vary.

One Windows app, three jobs

Voice Typing for the text you write all day. Meetings for the conversations you need to keep. Projects so neither gets lost.

  • Voice Typing

    Hold a key and speak — clean text appears in whichever Windows app you were writing in. Russian and English inside one sentence, cleanup before the text lands.

    How Voice Typing works

  • Meetings

    Capture your microphone and the system audio of a call, follow a live transcript with speakers and optional translation, and keep the recording with the notes.

    How Meetings work

  • Projects

    Every saved meeting keeps its recording, transcript, translation, notes and AI insights together — filed into Projects with search, instead of a flat history.

    How Projects work

Three reasons people switch

  • 1

    Up to 4x faster than typing

    Most people speak around 150 words a minute and type around 40. Dictating a long message takes a fraction of the time.

  • 2

    Two languages, one sentence

    Switch between languages mid-sentence without touching the keyboard layout. Xuula follows you.

  • 3

    Clean text, not a raw transcript

    Filler words go, punctuation appears, repeated starts disappear. What lands in the text field is what you meant to write.

Watch one sentence come together

This is what appears on screen while you speak, and what is left when you stop.

While you speak

Recording

What gets typed

Давай перенесём the design review на четверг — я отправлю the updated roadmap сегодня вечером.
An illustration of a text editor showing polished text after dictation.

A recorded demonstration. Recognition quality depends on your microphone, your accent and how clearly you speak.

так давай э-э перенесём the design review на четверг и я я отправлю the updated roadmap сегодня вечеромДавай перенесём the design review на четверг — я отправлю the updated roadmap сегодня вечером. (2)

From live words to a filed conversation

A meeting runs through four visible states — you always know what is recorded, what is translated and where it will be saved.

An illustration of the live Meeting Workspace: a speaker-separated transcript with a translation beside it, a recording indicator, a microphone mute switch, and the Project the meeting will be saved to.
  1. 1

    Live transcript

    Words appear progressively while people speak, separated by speaker. Your microphone has its own mute switch.

  2. 2

    Original and translation

    Optional live translation runs beside the original — in the workspace and in movable always-on-top caption windows.

  3. 3

    Saving to Project

    Pick the destination Project in the live window. Stopping shows an explicit saving state — no silent writes, no lost meetings.

  4. 4

    Summary, key points, to-dos, quotes

    After saving, generate AI insights on demand — one tab or all four. They stay with the conversation. AI output can contain mistakes; review before you rely on it.

How Meetings work

Replay the moment, not the whole hour

A saved conversation plays its recording right inside the transcript. The spoken word is highlighted as playback advances, and clicking any word seeks the audio to that exact moment — checking a quote takes seconds.

How Projects work

An illustration of the conversation library: a sidebar with Inbox and Projects, a saved conversation playing its recording with the current word highlighted, and tabs for the AI insights.

The arithmetic behind “4x”

A typical speaking pace is about 150 words a minute. A typical typing pace is about 40. That is a ratio of 3.75 — which is where “up to 4x faster” comes from. It is an illustrative comparison, not a guaranteed result, and it does not include the time you spend thinking.

180
Speaking150 words per minute
1.2 min
Typing40 words per minute
4.5 min
Differenceabout 3.3 min

At 1,500 words a day, that difference adds up to roughly 10 hours a month.

Before and after cleanup

Speech is full of false starts. Xuula removes them instead of transcribing them.

What you actually said

so um i think we should uh we should probably move the deadline because like the api the api is not ready yet

What Xuula types

I think we should move the deadline, because the API is not ready yet.

  • Removed “um”, “uh”, “like”
  • Removed the repeated start
  • Added punctuation

Cleanup runs on the text of your dictation. You can turn it off in settings.

Built for people who mix languages

Product names, technical terms and quotes rarely arrive in the language you started the sentence in. Xuula recognises the switch without you changing anything — no keyboard layout, no menu, no second app.

Russian and English are the launch focus. The product is built to add more languages, and your account can already store several preferred languages.

  • Example 1

    Отправь мне the deployment checklist до стендапа.

    Russian sentence, English product term

  • Example 2

    Let me check с командой and get back to you.

    English sentence, Russian clause

  • Example 3

    Нужно обновить the CI pipeline, там упал integration test.

    Technical terms kept in English

No keyboard layout switching. No “dictation language” dropdown.

Three steps

  1. 1

    F8Press F8

    Put the cursor wherever you want the text. Hold the key, or press it once.

  2. 2

    Speak

    Say it the way you would say it out loud. Switch languages mid-sentence if that is how you think.

  3. 3

    Get polished text

    Release the key. Cleaned-up text appears in the field, ready to send.

What you get

Small things that decide whether voice typing survives past the first week.

  • Works in any Windows text field

    Email, chat, browser forms, code editors, ticket trackers, the search box. If you can type into it, you can talk into it.

    No per-app plugins and no browser extension.

  • Personal dictionary

    Teach it the names, brands and technical terms that generic models get wrong — your colleagues, your product, your stack.

    Add spelling variants so a misheard word maps to the right one.

  • Writing styles

    Keep a message short and direct, or make a report read formally. Set a default and switch when you need to.

    Styles apply to cleanup, not to what you said.

  • Automatic cleanup

    Grammar fixes, punctuation, capitalisation, filler removal and repeated-word removal, applied as the text is inserted.

    Turn any of it off if you would rather have the raw transcript.

  • Stable words while you speak

    Words settle as you go instead of rewriting themselves. You can watch the sentence build without losing your place.

    The final cleanup runs when you stop.

  • Dictation history

    Everything you dictated recently, kept on your PC, so you can recover a message you sent into the wrong window.

    Stored locally. Nothing is synced to our servers in this release.

  • One key, two modes

    Hold F8 to talk, or press once to start and again to stop. Pick whichever suits how you work.

    The hotkey is configurable.

  • You always know when it is on

    A clear on-screen indicator shows when the microphone is live, and it disappears the moment recording stops.

    Optional launch at Windows startup.

Where your words go

  • On your PC

    Hotkeys, text insertion, dictation history, meeting recordings, transcripts, notes and Projects. None of it is uploaded to Xuula servers in this release.

  • Speech recognition

    Audio is streamed to the speech provider only while recording is active — dictation, or a meeting you started.

  • Optional AI

    Transcript text is sent for cleanup or meeting insights only when the feature is on or you ask for it. Never the audio.

  • Your account

    Email, subscription state and a word count for the free allowance. Not your transcripts, not your audio.

Read the full data-flow description

Simple pricing

Start free. Upgrade when the weekly allowance stops being enough.

Free

$0forever

Enough to find out whether voice typing fits how you work.

  • 2,000 words every week
  • Mixed-language dictation
  • Works in any Windows text field
  • Basic smart cleanup
  • Dictation history stored on your PC
  • Up to 2 authorised devices
Create a free account

No credit card required

Most chosen

Pro

$12per month

billed monthly

For people who dictate most of what they write.

  • Unlimited dictation, subject to fair use
  • Full smart cleanup
  • Personal dictionary
  • Writing styles and custom preferences
  • Priority processing
  • Priority support
  • Up to 5 authorised devices

Compare plans in full

Common questions

Yes. That is the case Xuula is built around. You do not select a language beforehand or change your keyboard layout — you speak, and the recognition follows the switch. Russian and English are the launch focus, and your account can store several preferred languages.

Read the full help centre

Try it on your next long message

Install it, press F8, and see what a week of not typing feels like. The free plan does not ask for a card.

2,000 words free every week. No credit card.