Frequently asked questions
Is Verenu actually free?
Yes. Verenu itself is free and open source under MIT. There's no account, no subscription, and no trial period. You'll pay whatever your chosen AI provider charges. That's often nothing, since Groq's transcription tier is free, and Google's has a free limited tier too.
What does "bring your own key" (BYOK) mean?
Verenu doesn't run its own transcription or cleanup servers. You add an API key for Groq, OpenAI, Google, or AssemblyAI for transcription, and Verenu sends requests directly from your machine to that provider. AssemblyAI handles transcription only.
Does dictation work offline?
Yes, if you choose local models. Select a local transcription model plus a local cleanup model (or turn cleanup off) in Settings -> Models, and dictation runs fully on-device after the model files download. Local models are available on Windows and Apple Silicon Macs, Intel Macs and Linux aren't supported yet. If you use a cloud provider instead (Groq, OpenAI, Google, or AssemblyAI), that step needs a connection. See local AI dictation for the full breakdown.
What data leaves my device?
Only what your dictation needs. Recorded audio goes to your transcription provider, the raw transcript goes to your cleanup provider, plus the vocabulary, snippets, and formatting instructions from the active context group. Everything else, including settings, history, your context groups, and API keys, stays on your device. Full breakdown on the Privacy page.
Which platforms are supported?
Windows 10/11 and macOS (Apple Silicon & Intel) today. There's no Linux or mobile build yet. See the Roadmap for what's actively being worked on.
Why not Electron?
Verenu is built on Tauri instead. There is no bundled Chromium, idle RAM stays lower (roughly 200 MB), and it fits a background dictation tool better than a browser-shaped desktop app.
How is this different from Wispr Flow or SuperWhisper?
Those are closed-source, subscription products with their own backends. Verenu is open source, free, and BYOK, so your audio and text go straight to the provider you chose, never through a Verenu-run server. See the Wispr Flow comparison for a detailed side-by-side.
Can I read or modify the code?
Yes, the full source is on GitHub under the MIT license. See Contributing if you want to send a patch.