Fluentry
Fluentry is a free, open-source Linux dictation application: users hold a key, speak, and have locally processed speech typed into the application they are using. Its speech recognition runs on the user's own machine so dictation works offline and audio is not uploaded unless the user explicitly enables AI cleanup and points it at a provider.
At a Glance
- Linux desktop users on Debian, Ubuntu, and other distributions
- Wayland and X11 users
- Developers and technical users who dictate prompts, code, or terminal commands
- Privacy-conscious individuals who want local speech recognition
- +1 more
AI Tools by Fluentry
(1)
Fluentry
Open Source Linux Dictation App
Discussions
No discussions yet
Be the first to start a discussion about Fluentry
Latest News
Fluentry 1.0.4: Debian 13 installation fixes and start-at-login by default
Project adds packaged XDG autostart and bumps to 1.0.3
Fluentry 1.0.2: GNOME focus extension is enabled automatically for terminal and Codex paste support
Fluentry 1.0.1 adds an eleven-language interface and Arabic right-to-left layout
Products & Services
Open-source Linux desktop dictation app distributed as a Debian/Ubuntu package and installable with pip. It captures speech, transcribes locally, and inserts the result into the active application; it also provides a tray app, recording overlay, searchable local history, custom dictionary, spoken punctuation, formatting rules, optional AI cleanup, a command-line transcriber, and an optional loopback HTTP API.
Public GitHub source repository for the Python/PySide6 application, licensed GPL-3.0-or-later.
Market Position
Fluentry positions itself as a local-first, free/open-source Linux desktop dictation tool with explicit Wayland support, rather than a cloud transcription service. Comparable Linux speech-to-text and voice-control options identified in broader research include Speech Note, Talon, nerd-dictation/Vosk, whisper.cpp, and Spokenly; Fluentry differentiates through its native desktop workflow, local model execution, libei-based Wayland insertion, custom dictionary and spoken formatting pipeline, and optional local LLM cleanup.
Leadership
Founders
Denis Rosa
Fluentry's GitHub owner and primary release author. His public profile lists Munich and identifies him as Director of Developer Ecosystem and Experience at Couchbase; Devmio describes him as a Couchbase Developer Advocate, software engineer with 14+ years of experience, and a speaker/developer experienced with Java, Python, Scala, JavaScript, machine learning, big data, and microservices.
Executive Team
Denis Rosa
Creator, GitHub owner, and primary maintainer/release author
Director of Developer Ecosystem and Experience at Couchbase and a Couchbase Developer Advocate; public speaker and software engineer with 14+ years of experience across Java, Python, Scala, JavaScript, machine learning, big data, and microservices.
Founding Story
Fluentry began as a Linux reimplementation of the FluidVoice dictation application. The project sought to reproduce the useful behavior, settings schema, and backup format on Linux using native Linux interfaces; its NOTICE says no source from FluidVoice is included and that Fluentry is written in Python against Linux interfaces.
Business Model
Revenue Model
Free and open-source software distributed under GPL-3.0-or-later; the official site presents it as free and does not describe subscriptions, usage billing, advertising, or a paid service.
Pricing Tiers
The website describes Fluentry as free and open source; installation is offered through a Debian package or pip.
Target Markets
- Linux desktop users on Debian, Ubuntu, and other distributions
- Wayland and X11 users
- Developers and technical users who dictate prompts, code, or terminal commands
- Privacy-conscious individuals who want local speech recognition
- Multilingual users needing an eleven-language interface and speech engines covering many languages
- Hands-free dictation into editors, browsers, chat applications, terminals, and other Linux desktop applications
- Writing long prompts or messages without typing
- Privacy-sensitive offline transcription
- Developers dictating into terminals and coding tools such as Codex
- Users who need multilingual Linux dictation and Wayland compatibility
- Transcribing recorded WAV files from the command line