Speech to text software for Windows 10 PCs
Press the hotkey. Get the reply, ready to send.
Installs on Windows 10 or later, no Windows 11 required. Local models work fully offline, once downloaded.
Open source · 80,000+ downloads
What stalls dictation on Windows 10
Windows 10 dictation stalled. Yours doesn't have to.
You need something that installs on the OS you actually have, keeps working without an internet connection, and gets your words right the first time.
Win+H needs the internet, every time.
Windows 10's built-in dictation requires a live connection to work at all, and it stops responding for some people after a Windows update, with no fix in sight. OpenWhispr's local models run entirely on your device, so a dropped connection never stops you mid-sentence.
Voice access never came to your PC.
Microsoft's newer voice tools are gated to Windows 11 version 22H2 and later; Windows 10 keeps the deprecated Windows Speech Recognition, and it isn't getting better. OpenWhispr installs on Windows 10 or later. No upgrade required.
It typed "background noise" into an email.
Generic recognition mishears background sound as words and turns punctuation into a spoken chore. OpenWhispr's custom dictionary learns your names and terms, and local speech models built for accuracy do the listening.
The professional option you remember is gone.
The consumer edition of Dragon was discontinued in 2023, with no replacement at that price. OpenWhispr is free, with unlimited local dictation for as long as you use it. No card required.
What you get
Dictation that doesn't care which Windows you're on.
Cloud is the easy choice for an older CPU; local is the choice when you want zero reliance on a connection. Both run on the Windows 10 you have today.
Local processing that never needs a connection
Download a Whisper or Parakeet model once, and every dictation after that runs on your own device. No internet, no server, and it keeps working through the day the Wi-Fi drops.
Cloud transcription with nothing to download
Sign in and dictate. OpenWhispr Cloud transcribes on our servers for the fastest, most consistent results, with no model to download and no strain on an older machine.
Dictate wherever your cursor already is
Press a hotkey and speak. The text is typed into Word, Outlook, Edge, Teams, Notepad or any app that accepts typed text. No per-app setup needed.
Internet needed in local mode
None
once the model is downloaded, dictation works fully offline
Cost for unlimited local dictation
$0
free forever, no Windows 11 upgrade needed
Setup takes
5 min
from download to your first dictated line on Windows 10
How it works
Installed on the Windows 10 you already have.
Start free. Three simple steps and you're dictating today.
Download for free.
Windows 10 or later, 64-bit. No Windows 11, and nothing extra to grant beyond your microphone.
Dictate the note.
Press the hotkey and speak. Choose Cloud for speed, or a local model if you'd rather skip the connection entirely.
Review, paste, done.
Dictation cleanup tidies filler words, grammar and punctuation before it lands where your cursor is, so a quick reply is ready to send. Edit if you need to, and you're done.
No convincing needed
10 reasons Windows 10 users choose OpenWhispr
Not a workaround for an OS Microsoft moved past. A dictation tool built to run on it.
Runs on Windows 10, not just Windows 11
The minimum requirement is Windows 10 or later, 64-bit. No Windows 11 gate, no new hardware to buy.
Works with no internet connection at all
Local models process entirely on your device once downloaded, unlike Windows' built-in dictation, which needs a live connection every time.
The published requirement has no expiry date
Microsoft's own Office security updates on Windows 10 are set to run through 10 October 2028; Edge and WebView2 keep updating on Windows 10 22H2 until at least that same date. OpenWhispr's requirement is simply Windows 10 or later, with no end date attached to it.
Free, with no card and no expiry
Unlimited local dictation on the Free plan, forever. No trial clock, no card required to start.
Dictate into whatever you're already using
One hotkey types clean text into Word, Outlook, Edge, Teams, Notepad, or any app with a text field. Nothing to configure per app.
Hold to talk or tap to talk, your choice
Windows supports both activation modes for dictation: hold the key while you speak, or press once to start and again to stop. Built-in dictation forces an all-or-nothing switch between speaking and typing that breaks your rhythm; picking the mode that fits you means it doesn't.
Teach it the words it keeps getting wrong
A custom dictionary learns names, product codes and jargon, so the terms you use every day stop coming out mangled.
GPU acceleration for local models, one click away
NVIDIA, AMD and Intel GPUs get a one-click acceleration download from Settings. On older hardware, pick a smaller local model, or skip local entirely and use Cloud, which asks very little of your computer.
Say what you want changed, not just what you said
Select the text, press the voice agent hotkey (set it once in Settings), and say what to change: "make this more formal", "fix the typos". The rewritten version lands in place, not your raw words.
Open source, so IT can check for themselves
Every line is public on GitHub. Before it goes on a work PC, someone can read exactly what happens to the audio.
Before you download
The questions Windows 10 users ask first.
Straight answers on compatibility, cost and accuracy. If yours isn't here, ask us in Discord. We reply within 24 hours.
Does OpenWhispr actually run on Windows 10?
Do I need an internet connection to dictate?
Will it run well on an older Windows 10 PC?
How is this different from Windows' built-in dictation?
What does it cost?
Can it learn my own words and terms?
What users say
What people actually said, unedited.
From our Discord, GitHub and inboxes: people who kept an older machine going and wanted the dictation to keep up.
0
Files stored on a server
In local mode the audio never leaves your device. Nothing is stored on a server, and nothing is sent to a transcriber.
Started using OpenWhispr with my local models and hardware. I have to say, this is better than typing.
I switched to voice input for my Claude Code chats. Even when I stumble and rephrase mid-sentence, the transcription comes out clean and polished.
I needed a local model for when I'm travelling or have no WiFi, and this perfectly solves it. The UI and performance are great.
Pricing
Free forever.
Upgrade when you want more.
FreePopular $0No card details needed | Pro $6.67/user/mo$80/user billed annually Start with Pro | Business $13.33/user/mo$160/user billed annually Start with Business | |
|---|---|---|---|
| Best for | Windows 10 users who want free, unlimited local dictation, no Windows 11 or internet required | Windows 10 users who want unlimited Cloud transcription plus mobile, sync, and API/MCP access | Teams keeping older Windows 10 machines productive, with unlimited meeting recordings and chat over data |
| Main reason to choose | Dragon Home, the affordable paid option, is gone from sale. OpenWhispr is free, and stays free | Cloud transcription that doesn't ask anything of an older CPU | Unlimited meeting recordings plus chat over your data |
| Offline speech-to-textOffline speech-to-text runs entirely on your own device, it works without internet connection, and it's unlimited and free. | Included | Included | Included |
| Cloud speech-to-textCloud speech-to-text sends audio to OpenWhispr Cloud for the highest accuracy and speed, with no local models to download. | 2,000 words/week via OpenWhispr Cloud, or unlimited using your own API keys | Unlimited OpenWhispr Cloud speech-to-text | Unlimited OpenWhispr Cloud speech-to-text |
| Meeting recordings | 5 hours/month | 20 hours/month | Unlimited |
| Privacy & data handling | Offline audio stays on your device. Cloud audio is sent only to transcribe it, then never stored. | Offline audio stays on your device. Cloud audio is sent only to transcribe it, then never stored. | Offline audio stays on your device. Cloud audio is sent only to transcribe it, then never stored. |
| Setup experience | Hands-on: install app, choose local models, optionally add API keys | Easy: install app, upgrade, use unlimited cloud speech-to-text | Moderate: configure users, Agent mode, and data workflows |
| Advanced features |
|
|
|
| Team management | None | None | Basic team / workflow use |
| Support | Community support | Email support | Priority support |
Need SSO, compliance features, admin controls, or dedicated support for a larger team?
We also offer a custom Enterprise plan with managed onboarding and tailored features.
Talk to us
Built in the open
Some machines aren't getting upgraded. We built for those too.
OpenWhispr started with a simple objection: dictation software shouldn't stop working the day your OS falls out of support. Windows 10 is still Windows, so dictation, Cloud and local transcription work the same as they do everywhere else we ship.
We build in public. Every release and every line of code is on GitHub, where 5,000+ engineers have starred, forked and audited the project. When we say local dictation runs on your device, that's not a promise in a privacy policy. It's a property of the code, and anyone still running Windows 10 can verify it themselves.
And if anything breaks, you talk to the people who built it, in Discord.
Free · open source · local-first
Your words, typed on the Windows 10 PC you already have.
Download OpenWhispr and your next dictation lands as clean text, offline if you want it. No Windows 11 required.
Loading...Takes 5 minutes to set up. If it doesn't work, we'll fix it with you in Discord. Free forever on Mac, Windows & Linux.