Hold a key, say your sentence, release. The text lands wherever your cursor is — Telegram, LinkedIn, Word, your terminal. And when a meeting starts, the same app records it and writes the summary.
One free hour of recognition per month · email only, no keys · works in every app
Hold Right Alt and talk. Release — the recognized text is pasted at your cursor. Filler words are cleaned up, paragraphs are kept, a custom vocabulary protects names and brands from being mangled.
No bot joins your call. Vymova records your microphone and the system audio as two separate channels — Zoom, Meet, Teams, any call, even an offline conversation — and delivers a live transcript plus an AI summary.
Interface, text cleanup and vocabulary are built for Ukrainian — not translated as an afterthought. English, Czech, German and more are supported too.
Dictation history and meeting transcripts live only on your computer. Cloud sync exists, but it is off by default and only ever syncs meetings — never dictations.
Sound is captured from your own sound card, so every meeting gets recorded — not just the ones where an "AI notetaker" was allowed in.
Words appear seconds after they are spoken and correct themselves as the phrase completes. A clean, verified transcript is produced after the meeting ends.
ChatGPT, Claude, Cursor, Slack, Gmail, Jira — no per-app integrations, because pasting at the cursor needs none.
Word, Markdown, JSON export. A webhook can deliver every finished transcript with its summary straight into your CRM or task tracker.
One file. Desktop shortcut and launch on startup — optional.
One field in Settings. Recognition keys are not needed — your first hour each month is on us.
The text appears where your cursor is. Rebind the key to anything you like.
Yes. Install, enter your email, and dictate: one hour of audio per month is recognized on our account. No card, no third-party sign-ups. When the hour runs out, plug in your own free Groq key — then there are no limits from us at all.
Recorded snippets are sent to the speech-recognition service and come back as text. On the free plan it passes through our proxy and is never stored; with your own key it goes directly, bypassing us entirely. History and transcripts stay on your computer.
Yes — and in Teams, Slack huddles, browser calls and offline conversations. Audio is captured from the sound card, so the calling app does not matter.
We are working on a code-signing certificate. A SHA-256 checksum is published next to the installer so you can verify the download; the details are on the download page.