SpeakUp Is Coming to Windows

Why Windows?

Mac users have been using SpeakUp to dictate without the cloud since day one. Now we're bringing the same experience to Windows — because privacy shouldn't depend on which computer you use.

What You Get

On-Device Whisper AI

Your voice is transcribed entirely on your PC. Nothing is sent to the cloud. Nothing is stored. Nothing is shared. The AI model runs locally, using your GPU for fast, accurate transcription.

Works in Every App

Press a hotkey, speak, press again. Your words appear wherever your cursor is — Word, Outlook, Slack, your browser, any app. No plugins, no extensions, no setup.

GPU-Accelerated

SpeakUp for Windows uses NVIDIA CUDA and DirectML to deliver fast transcription on modern GPUs. Intel, AMD, and NVIDIA — all supported. No GPU? It runs on CPU too.

German + English from Day One

Built by a team in Berlin, SpeakUp has first-class German and English support — with more languages coming. If you dictate in German, you'll feel the difference compared to tools that treat it as an afterthought.

No Subscription

One price, forever. No monthly fees. No word limits. No usage caps. No "free tier" that nags you to upgrade. You buy it, you own it.

The Dragon Question

Nuance discontinued Dragon Home. Dragon Professional costs $699 and hasn't seen a meaningful update in years. Microsoft is folding the technology into Copilot+ PCs — but only for English, and only on new hardware with NPUs.

If you're a Dragon user looking for what comes next: we built SpeakUp for you. Modern AI, on-device processing, a price that makes sense, and active development.

How It's Built

SpeakUp for Windows is built with Tauri and whisper-rs — a Rust wrapper around OpenAI's whisper.cpp. The result is a lightweight native app that uses your system's GPU directly, with no Electron bloat and no web runtime overhead.

*Fluid Dictation is on-device but requires a Copilot+ PC with NPU hardware.
**Fluid Dictation is English only.