Say it. It's written.
Emails, documents, messages, code. Hold AltGr, speak, then release. Your words are pasted into the app you're using.
Works with text fields that accept pasted text.
Your voice.
Already written.
Free offline voice dictation for Windows.
Speak into your apps. Keep your thoughts on your PC.
Windows 10 / 11 · Free & open source
“The first draft is ready. Let's review it together tomorrow.”
Illustrative demo · No microphone access
Small on your screen.
Part of your flow.
One shortcut between a thought and something useful. WritHer stays close, then gets out of the way.
To the team
A little less typing.
The first draft is ready. Let's review it together tomorrow.
Written where you work.Catch the thought.
“Save a note: try a shorter introduction.” Speak to your local assistant with Ctrl + Alt + R and keep the idea for later.
Assistant mode needs a local language model.
Leave yourself a reminder.
“Remind me to review the draft tomorrow at ten.” Turn the next thing on your mind into a reminder or an appointment.
Notes and schedules stay in your local database.
Your voice.
Your computer.
Speech recognition runs locally with faster-whisper. Download the speech model once, then dictate without an internet connection.
For the assistant, connect Ollama or an OpenAI-compatible local server. Keep that server on your own machine for a fully local setup.
No accountStart without signing up.
No word quotaFree software, under the MIT license.
Open to inspectionRead the code. See how it works.
See it on a real desktop.
From holding a key to capturing your next idea. Watch WritHer in action.
The video may show an earlier version of the interface.
Make WritHer yours.
A personal tool. A shared project.
Everyone is welcome to make it better.
Start with your own way of working.
Your shortcuts. Your vocabulary. Your language. Adapt WritHer to how you think and speak. The code is open, too: change the widget, try a different idea, or build something we haven't imagined yet.
Explore the settingsBuild somethingImprove the code, design a detail, explore a new platform.Bring your perspectiveShare an idea, report a bug, tell us what could work better.Help it speak to more peopleImprove a translation or help someone get started.
You don't need to be a developer. Your first idea, your first bug report, your first contribution: they all belong here.
Join us on GitHubA few minutes to get started.
Start with dictation. Add the local assistant when you're ready.
Download & extract
Get the Windows ZIP from GitHub Releases. Extract the folder and run
WritHer.exe.Let the model load
The speech model downloads on first launch. You'll need an internet connection for this step.
Hold a key & speak
Focus a text field, hold AltGr, speak, then release. Change shortcuts or recording mode in Settings.
A few good questions.
Read the WritHer FAQIs WritHer really free?
Yes. WritHer is open source under the MIT license, with no subscription or paid dictation tier. You can use it, inspect it, modify it and contribute.
Do I need Ollama for dictation?
No. Speech-to-text works on its own with faster-whisper. Only the voice assistant needs a language model, through Ollama or another supported local server.
Does it work offline?
Dictation works offline after the speech model has downloaded. The assistant also works offline when its server and model run locally. Choosing a remote server changes where assistant requests are sent.
Can I use it on macOS or Linux?
WritHer currently supports Windows 10 and 11. Contributions towards other platforms are welcome. See the contribution guide.
A free, offline alternative to Wispr Flow.
If you're looking for speech-to-text on your own Windows PC, WritHer gives you local dictation and an open-source voice assistant. Compare the approaches and choose what fits your workflow.
WritHer vs Wispr Flow
Your next idea
starts with your voice.
Free to use. Open to everyone.