100% Offline Speech Recognition

Speak Your Mind, Watch It Type

Transform your voice into perfectly formatted text with best-in-class AI accuracy. PeekoType works 100% offline, so your voice never leaves your device. GDPR-compliant by design. Buy once, use forever. No subscriptions — the optional AI tools on mobile are pay-as-you-go, billed by OpenAI for what you use.

90-95% Accuracy
100+ Languages
100% Offline

Dragon accuracy without Dragon prices

Dragon Professional $699
vs

Same offline capability. Industry-leading AI accuracy. 94% less cost.

UK & EU GDPR-Compliant by Design

Your voice never leaves your computer.
No cloud. No transfers. No paperwork.

Most cloud dictation apps stream your audio to US servers and then point at UK-US data transfer frameworks to claim GDPR compliance. PeekoType doesn't need any of that, because nothing ever leaves your device. All speech recognition runs locally on your own CPU. No account, no telemetry, no processor agreements.

No international transfers
Audio is processed on your PC. Nothing crosses a border, so the UK–US Data Bridge / EU–US DPF simply don't apply.
No data processor
No third party ever sees your audio or transcripts. There is no processor to put under a DPA, so you are the controller of your own data.
No telemetry, no account
No analytics phoning home, no sign-up, no email collection inside the app. Even your licence check is performed locally.
HIPAA & legal-friendly
Suitable for healthcare notes, legal dictation, journalism and any work where cloud transcription is off the table.
Portable mode
Run the whole app from a USB stick. No install, no registry entries, no admin rights, nothing left behind on the host PC.

Want the technical detail? Read why PeekoType is GDPR-compliant by design →

Why Choose PeekoType

Powerful features that make voice typing effortless

Advanced Offline Speech Recognition

Industry-leading on-device AI delivers exceptional accuracy across 100 languages, handles accents naturally, and adds punctuation automatically. Read more →

100% Offline & GDPR-Compliant

Your voice never leaves your computer. No cloud. No telemetry. No data collection. GDPR-compliant by design and HIPAA-friendly for healthcare professionals. Read more →

Works in ANY App

Word, Notepad, Slack, Email, browsers, code editors. PeekoType types into any text field, anywhere. Read more →

90-95% Accuracy

Near Dragon-level accuracy powered by advanced offline AI. Handles technical terms, accents, and background noise. Read more →

Auto Punctuation

Smart punctuation added automatically. Capitalisation, full stops, commas, all handled for you. Read more →

£

Buy Once, Own Forever

No subscriptions. No monthly fees. One payment and it's yours for life. No "you've reached your limit" messages. Read more →

Filler Word Removal

Strips "um", "uh", "hmm" and other filler sounds on the fly. Your transcripts come out clean and ready to share. Read more →

1. 2. 3.

Auto-Format Lists

Say "make a list" and PeekoType numbers your items automatically. Perfect for shopping, agendas and meeting notes. Read more →

Voice Commands

Say "period", "comma", "new line" or "delete" to control punctuation and edits hands-free, mid-sentence. Read more →

Natural-Language Edit Commands

"Delete that last sentence." "Scratch that." "Undo." Correct what you just dictated without touching the mouse or keyboard. Read more →

Custom Vocabulary

Teach PeekoType the names, jargon and acronyms it keeps mishearing. Recognition improves for the words specific to you. Read more →

Voice-Triggered Text Snippets

Say "insert signature" and a saved block of text — a sign-off, a disclaimer, boilerplate — is typed in full, instantly. Read more →

Searchable Transcript History

Every dictation is logged locally and instantly searchable. Nothing vanishes when the next sentence overwrites the screen. Read more →

On-Device Translation

Translate any past transcript into another language, fully offline, with the original kept alongside it. Read more →

Read-Aloud Voices

Hear any transcript — or its translation — read back in one of 49 languages. Built for proofing, and for dyslexia- and ADHD-friendly workflows. Read more →

Live Captions, No Bot

Subtitle any Teams, Zoom or Meet call in a floating overlay. Nothing joins the meeting and nobody else can tell — it reads the audio already playing on your device. Read more →

Speaker Detection

Transcribe an interview or a meeting and get speaker labels, so it reads as a conversation instead of one unbroken block of text. Included in the one price. Read more →

TXT, SRT, VTT, DOCX & PDF

Batch-transcribe whole folders of audio and video, and export as plain text, timed subtitles, a Word document or a PDF — saved next to the original. Included in the one price. Read more →

Light & Dark Themes

Switch themes with one click, at any time. Everything re-tints instantly — including windows you already have open — so nothing you were working on is lost. Read more →

Five Model Sizes

Choose your own trade-off between speed and accuracy. Tiny flies on a decade-old laptop; Large is the most accurate of all if you have a GPU. Read more →

Portable Mode

Run PeekoType from a USB stick. Settings, vocabulary, snippets and history all live in one folder, leaving nothing behind on the host computer. Read more →

Built for People Who Value Their Time

Writers & Authors

Write 3x faster. Dictate novels, articles, and blog posts hands-free. Beat typing fatigue.

Professionals

Clear your inbox by speaking. Emails, reports, and documents at the speed of thought.

Students

Never miss a lecture. Capture notes and transcribe study materials effortlessly.

Privacy-Conscious

100% offline means 100% private. Your words stay on your computer. Always.

Healthcare Workers

HIPAA-friendly. Dictate patient notes without cloud privacy concerns. Fully local.

Journalists & Interviewers

Pro's speaker detection labels who said what in an interview or panel — automatically, offline.

Three Steps to Effortless Typing

Set your mic, pick your model, choose a language. Then just press F9.

Want the detail? The full user guide walks through every feature and button in the app.

PeekoType with the Microphone dropdown open, listing the input devices Windows has available
Step 1

Pick your microphone

Every input device Windows knows about is in the dropdown. Built-in laptop mic, headset, USB studio mic, they all work.

The PeekoType Dictate tab: record button, Smart output toggles and the Last transcription panel
Step 2

Pick your model size

Tiny is instant on any laptop. Large is the most accurate on a GPU. Small or Medium is the sweet spot for most people.

PeekoType with the Language dropdown open, showing English, Finnish, French, German and more
Step 3

Pick your language

100 languages supported. English, Spanish, French, German, Mandarin, Hindi, pick yours from the dropdown.

Press F9 and dictate

Click into any text field, hit F9 (or the record button), speak. The text appears where your cursor is. Done.

A copy is kept in the app's Last transcription panel too, so nothing is lost if it lands in the wrong window.

See the full walkthrough →

Live Captions for Teams, Zoom & Any Video Call

Real-time captions while you're on a call — without a bot joining the meeting, and without your audio ever leaving your PC. Read the full guide →

No bot joins the call

PeekoType listens to your PC's own audio directly — nobody on the call sees "PeekoType has joined," because nothing joins.

Works with any call app

Teams, Zoom, Google Meet, Slack huddles, Discord — if it plays audio through Windows, PeekoType can caption it.

49 languages

Live captions in 49 languages, using the same on-device engine as everything else — no cloud round-trip.

Nothing leaves your PC

Unlike bot-based meeting tools, your call audio is never uploaded anywhere to generate captions.

Included

Point It at a Recording Instead of a Microphone

Drop in an audio or video file — a meeting, an interview, a lecture — and get a full, exportable transcript. Still 100% offline. Read the full guide →

The PeekoType Transcribe tab: a drop zone for audio and video files, the transcription queue, and settings for output format, speaker detection and translation
Drop in files or a whole folder, choose your export format, and press Transcribe.
Batch transcription

Drag in files, or a whole folder, and transcribe them in one queued run.

Five export formats

TXT, SRT and VTT subtitles, DOCX, or a print-ready PDF — pick the format the transcript is going to.

Speaker detection

Automatically labels who's speaking in a multi-person recording — meetings, interviews, panels.

Translated transcript exports

Export a file transcript translated into another language, timestamps intact, alongside the original.

Simple, One-Time Pricing

No subscriptions. No hidden fees. Pay once, own forever.

How PeekoType Compares

Windows dictation tools, side by side. Prices and features are as published by each vendor, checked September 2026.

Product Price Offline Dictation File transcription Live captions Read aloud
PeekoType £39 one-time Yes Yes Yes Yes Yes
DictaType $14.99 plus ~$0.006/min OpenAI API No Yes No No No
Whisperstream $29 one-time Yes Yes Yes No No
FluidVox $39 one-time or $2.99/mo Local plan Yes Yes No No
MacWhisper (Mac only) €59 one-time Yes Limited Yes Yes No
Superwhisper $8.49/mo, $84.99/yr or $249.99 lifetime Local models Yes Yes (not on Windows) No No
Wispr Flow $15/mo or $144/yr No Yes No No No
Dragon Professional v16 ~$699 perpetual Yes Yes Limited No Yes

Comparison compiled September 2026 from publicly listed pricing and features. Vendors change their products, so check before you buy.

PeekoType on your phone

The same private dictation, as a keyboard. Tap the microphone in any app and speak — your voice never leaves the phone.

PeekoType on Android is a keyboard, not an app you dictate into and copy out of. You enable it once in your phone's settings, and from then on it appears wherever you type — messages, email, notes, documents, a browser address bar. Tap the microphone on the keyboard bar, speak, and the words land directly in whatever you were typing into.

Speech is turned into text by the same Whisper models the Windows app uses, running on the phone itself. There is no account, no sign-up and no connection needed once the voice has downloaded, and it understands around 100 languages. On a plane or underground it works exactly as it does at home.

There are things the phone version does that the desktop one cannot: a floating microphone button that follows you between apps, and read-aloud that speaks text back to you in a natural voice, generated on the device. If you supply your own OpenAI key, it will also type the words out live as you speak them rather than waiting for you to stop.

Free, for good

Dictation with the Tiny and Base voices, in about 100 languages. Offline, no account, no time limit, and no cap on how much you use it.

Punctuation and editing by voice — "new paragraph", "scratch that" — are part of the free app too.

£4.99 once

Adds the Small model, transcribing recordings and videos, custom vocabulary, read aloud, the floating microphone, live typing and AI cleanup. One payment, not a subscription.

Live typing and AI cleanup also need an OpenAI key you create yourself and pay OpenAI for directly. We take nothing from that.

Download for Android

Android 8.0 or later, 64-bit phone. 34 MB. Because this comes from us rather than a store, your browser will ask you to allow the install once — that prompt is Android checking you meant to, not a warning about the app.

SHA-256 561f833a4a885273cab27def86045785525f44971de00a0afa7f4ac4b00a9aaf

This is a separate product from PeekoType for Windows above. Buying one does not include the other.

Ready to Type at the Speed of Thought?

Try PeekoType free for 14 days. No credit card required to start.

Start Your Free Trial