Spokenly for Linux

Linux has always had powerful speech recognition engines, but very few polished dictation apps. Spokenly closes that gap: press a hotkey, speak, and accurate text appears in whatever app you are using — browser, IDE, terminal, email, or chat.

You choose how transcription runs. Local Whisper and Parakeet models process everything on your machine, free forever, with full privacy and offline support. Or connect cloud engines like Deepgram and OpenAI — with your own API keys at no extra cost, or through a Spokenly Pro subscription.

After transcription, built-in AI text processing removes filler words, fixes punctuation, and reformats what you said — a rambling thought becomes a tidy email, a bug report, or a commit message with a single mode.

Spokenly is system-wide: it types where your cursor is. Dictate a pull request review in the browser, comments in your editor, a message in Slack or Discord, or a whole document in LibreOffice. If you can type there, you can dictate there.

A global hotkey starts and stops recording, and a small overlay shows that Spokenly is listening. There is no separate window to copy text out of — your words land directly in the app you are working in, on both X11 and Wayland.

Beyond live dictation, Spokenly converts pre-recorded audio and video to text. Drop in MP3, WAV, M4A, or MP4 files and get clean transcripts — the jobs that usually require a separate Linux transcription tool, like interviews, lectures, voice memos, and meeting recordings.

Spokenly for Linux ships in three formats: a deb package for Ubuntu, Debian, and Linux Mint, an rpm package for Fedora and other rpm-based distributions, and an AppImage that runs everywhere else, including Arch. If your machine runs a modern desktop, Spokenly runs on it.

Until now, dictation on Linux mostly meant do-it-yourself tooling. nerd-dictation is a capable command-line tool, but you assemble the pieces yourself: models, hotkeys, and text injection. Desktop tools like Speech Note focus on transcription in their own window rather than typing into every app. Open-source Handy runs on Linux and shares the local-first idea, but stays deliberately minimal.

The apps people actually love on macOS and Windows — Wispr Flow, Superwhisper, MacWhisper — never shipped Linux versions. Spokenly brings that class of dictation experience to the Linux desktop, with the same features as our Windows and macOS apps.