Hyprwhspr
Hyprwhspr is free, MIT-licensed local speech-to-text for Linux with system-wide dictation on Wayland and X11.
If your machine runs Linux and you have a halfway decent GPU, this is the most capable free dictation setup you'll find — six local backends, hot models, and zero telemetry. The flip side is deliberate: source builds only and configuration through files and CLI. Choose Hyprwhspr if you'd rather read your dictation tool's code than click through an installer; pass if you need a GUI settings panel or a one-click binary.
Verified 5d ago · liveness 69/100 · cite: rightaichoice.com/tools/hyprwhspr
- Linux developers who want to dictate code, commit messages, and AI prompts into any focused window
- Privacy-conscious users who need audio processed locally with no account and no telemetry
- Wayland users on Hyprland, Sway, or KDE who want a themed on-screen mic overlay
- Writers and journalists doing hands-free long-form drafting with pause-and-save
- Anyone who wants a graphical settings app instead of file and CLI configuration
- Users who need a precompiled binary — installation builds from source by design
- Linux newcomers uncomfortable cloning a repo and running install scripts
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Hyprwhspr if you want a one-click precompiled installer and a graphical settings app rather than building from source and editing config files.
Cloud streaming through Gemini, OpenAI, or ElevenLabs is billed by those providers against your own API key — Hyprwhspr itself costs nothing, but heavy streaming use lands on your provider invoice.
Hyprwhspr is MIT licensed and free with no subscription, telemetry, or monetization — the default local setup needs no API key or account. That puts it below paid Linux dictation tools and cloud dictation subscriptions on cost; the tradeoff is setup effort rather than dollars, since accuracy with cloud backends depends on your own Gemini, OpenAI, or ElevenLabs API key and its provider billing.
In short
Hyprwhspr — Hyprwhspr is free, MIT-licensed local speech-to-text for Linux with system-wide dictation on Wayland and X11. Best for Linux developers who want to dictate code, commit messages, and AI prompts into any focused window, Privacy-conscious users who need audio processed locally with no account and no telemetry, Wayland users on Hyprland, Sway, or KDE who want a themed on-screen mic overlay. Free to use.
What people actually say about Hyprwhspr — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
13 mentions across 2 sources (Hacker News, GitHub) · researched Jul 6, 2026.
Average across the 2 sources that answered — each source counts once, not each post.
- +Fast, accurate local transcription using cutting-edge models like Cohere Transcribe.
- +Privacy-first: all inference runs locally, no data leaves your machine.
- +Deep Wayland integration: visualizer overlay, Waybar status, per-app paste rules.
- +GPU acceleration auto-detected for NVIDIA, AMD, and Intel Vulkan.
- +Multiple recording modes: toggle, push-to-talk, continuous, long-form, auto.
- −Audio device changes require manually restarting the systemd service.
- −Terminal escape sequence fragments appear after paste injection in some setups.
- −Requires a GPU for optimal performance with large models.
- −No progress indicator or spinner for long transcription tasks.
- −First recording start may trigger pulseaudio timeout errors.
- • Requires a GPU for best performance; may need to purchase one.
- • Optional cloud API usage (Gemini, OpenAI) may incur costs from those providers.
Viability Score
How well maintained and how widely used is Hyprwhspr? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- System-wide speech-to-text on Linux that injects text into any focused application via wtype or ydotool
- Local engines: Parakeet TDT V3, Cohere Transcribe, Whisper family, faster-whisper, onnx-asr, Qwen3-ASR
- GPU acceleration with auto-detected NVIDIA CUDA and AMD/Intel Vulkan, CPU fallback through onnx-asr
- Models held hot in memory plus on-demand VRAM unload and instant reload without restarting the service
- Five recording modes: toggle, push-to-talk, auto tap-vs-hold, continuous with auto-flush on silence, long-form with
- Per-app paste rules with optional auto-submit on paste
- Audio ducking that lowers other audio or pauses your player during recording
- Themed mic visualizer overlay that auto-matches your Omarchy theme on Hyprland, Sway, and KDE
- Waybar live status tray and GNOME desktop notifications for idle, recording, processing, and error states
- Multilingual transcription with optional translate-to-English mode
- Text processing: word overrides, filler word removal, symbol replacements, custom prompts
- Opt-in WebSocket streaming in near realtime via Google Gemini, OpenAI, or ElevenLabs
- Self-healing recovery from suspend/resume, USB mic unplug, and keyboard hotplug, with mic-muted warnings
- Post-transcription hook pipes every result through a custom shell command before pasting
- Scriptable CLI for start/stop/toggle/cancel plus built-in evdev hotkeys
About Hyprwhspr
Hyprwhspr is a system-wide speech-to-text tool for Linux. You speak, it injects text into whatever buffer currently has focus — your editor, terminal, browser, chat app. It runs on Wayland and X11 across Hyprland, GNOME, KDE Plasma, and Sway, with GNOME/X11 on Ubuntu 24.04 called out as validated. The target user is a Linux developer or writer who wants dictation without shipping audio to a third party; local inference is the default path, and nothing is logged or tied to an account. The engine roster is the draw. It ships with Parakeet TDT V3, Cohere Transcribe, the full Whisper family, faster-whisper, onnx-asr, and Qwen3-ASR. GPU acceleration auto-detects NVIDIA CUDA or AMD/Intel Vulkan and drops back to CPU, where onnx-asr is treated as a real option rather than a consolation prize. Models stay hot in memory so you aren't paying a multi-second load per phrase, and you can unload one from VRAM on demand and reload it without restarting the service. Day-to-day features are dense for a free tool: five recording modes (toggle, push-to-talk, auto tap-vs-hold, continuous with auto-flush on silence, long-form with pause and save), audio ducking that steps other playback down or pauses your player, per-app paste rules with optional auto-submit via wtype or ydotool, a themed mic overlay that matches your Omarchy theme, a Waybar status tray, and multilingual transcription with optional translate-to-English. A post-transcription hook pipes each result through your own shell command before pasting. An opt-in WebSocket path streams near-realtime to Google Gemini, OpenAI, or ElevenLabs using credentials stored outside your config files. Installation is the honest caveat. There is no precompiled binary by design — the project argues a black box has no audit trail for something sitting on your microphone. You build from source via the AUR (yay -S hyprwhspr) or clone the repo and run install-deps.sh plus the setup script. Against Talon or a cloud dictation subscription,
Behind the Verdict
Hyprwhspr solves a problem Linux users have grumbled about for years: dictation that actually works system-wide and doesn't phone home. The default posture is local inference, and the privacy claim is specific — no account, no logging, audio stays put unless you explicitly configure a cloud backend with your own API key. Pick this when you dictate into many different apps and want one hotkey to rule them all. Per-app paste rules, optional auto-submit, and a post-transcription hook that pipes text through your own shell command cover the cases where a generic dictation app falls apart. Hook that into a commit-message formatter or a logging script and it stops being a dictation tool and starts being a pipeline component. The CLI (start/stop/toggle/cancel) plus built-in evdev hotkeys means you can wire it to sxhkd, Espanso, or your own scripts. We'd also reach for it on weaker hardware. Most local transcription tools treat CPU as the sad path. Hyprwhspr ships onnx-asr and calls it fast and accurate in its own right, which matters if you're on a laptop without a discrete GPU. Where it bites: there is no precompiled binary, and that's stated as philosophy, not backlog. You build from source — AUR on Arch, install-deps.sh plus setup on Debian, Ubuntu, Fedora, and openSUSE. Configuration lives in files and CLI flags. If you want a settings window with sliders, this isn't your tool, and no amount of community enthusiasm will change that this quarter. The closest alternative depends on what you're optimizing for. Talon offers more polished voice-command control and a paid commercial path; cloud dictation subscriptions trade privacy and a monthly fee for zero local setup. Hyprwhspr undercuts both on cost (free, MIT, no monetization) and beats them on backend breadth —
Researching Hyprwhspr? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Hyprwhspr actually fits — and what changes day-one when you adopt it.
You bind Super + Alt + D to toggle recording, keep Parakeet TDT V3 hot in memory, and dictate commit messages and prompts directly into your terminal and editor via wtype with per-app paste rules.
Outcome: Hands-free drafting in any active buffer without touching the keyboard, with transcription ready instantly because the model never unloads.
You keep the default local backend so audio never leaves the machine, use long-form mode with pause and save for drafting sessions, and route every finished transcription through a post-transcription shell hook that files it into your notes directory.
Outcome: Drafting stays entirely on your own hardware, and output lands in your pipeline automatically instead of being pasted and manually moved.
You dictate in your native language with translate-to-English mode enabled, and use continuous mode with auto-flush on silence for hands-free input during calls or note-taking.
Outcome: Spoken input arrives as English text in whatever field is focused, with no manual switching between languages.
Use Cases
- Dictate code, commit messages, and prompts hands-free into your IDE or terminal
- Transcribe meeting notes, journal entries, or email drafts into any text field
- Control your system with voice through custom scripts and hotkeys
- Translate spoken input to English in real time for multilingual work
- Automate text input in applications using continuous mode with auto-flush on silence
Models Under the Hood
as of 2026-10-17
Limitations
- Hyprwhspr is Linux-only, supporting Wayland and X11 across Hyprland, GNOME, KDE Plasma, and Sway (GNOME/X11 on Ubuntu 24.04 is the validated configuration).
- It is driven via CLI, hotkeys, and tray/overlay indicators, with configuration stored in files rather than a graphical settings app.
- GPU acceleration requires compatible NVIDIA CUDA or AMD/Intel Vulkan drivers, falling back to CPU via onnx-asr.
- Cloud streaming and provider features (Gemini, OpenAI, ElevenLabs) require you to supply your own credentials, and the REST API and WebSocket streaming are opt-in.
as of 2026-10-02
Verification history
We have re-verified Hyprwhspr 9 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 9 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Hyprwhspr tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free & Open Source (MIT)
$0
Ideal for
Linux developers, writers, and privacy-minded users on Wayland or X11 who are comfortable building from source and want local dictation with no account or subscription
What this tier adds
Starting tier — the only tier; everything ships here under MIT, including all local engines, five recording modes, GPU acceleration, and opt-in cloud streaming with your own API key
Where the pricing makes sense
The company stage and team size where Hyprwhspr's pricing actually pencils out — and where peers do it cheaper.
Hyprwhspr is MIT licensed and free with no subscription, telemetry, or monetization — the default local setup needs no API key or account. That puts it below paid Linux dictation tools and cloud dictation subscriptions on cost; the tradeoff is setup effort rather than dollars, since accuracy with cloud backends depends on your own Gemini, OpenAI, or ElevenLabs API key and its provider billing.
Setup time & first value
How long it actually takes to get something useful out of Hyprwhspr — broken out by persona, not the marketing-page minute.
Arch users get the fastest path: yay -S hyprwhspr installs the stable build and setup is minutes. Debian, Ubuntu, Fedora and openSUSE users clone the repo, run ./scripts/install-deps.sh, then ./bin/hyprwhspr setup — budget a longer block for dependencies and your first model download. Cloud streaming adds the time it takes to obtain and configure a Gemini, OpenAI, or ElevenLabs API key.
Switching to or from Hyprwhspr
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Talon: keep your existing voice-coding habits but move to local open-source models via the AUR package or the build-from-source setup script.
- →From cloud dictation subscriptions: switch the default backend to a local model so no audio leaves your machine, then optionally re-enable Gemini, OpenAI, or ElevenLabs streaming with your own key.
- →From whisper.cpp scripts: replace per-invocation shell-outs with Hyprwhspr's hot-in-memory models and add per-app paste rules plus a post-transcription hook.
- →From desktop environment dictation toggles: map Hyprwhspr's CLI start/stop/toggle/cancel or evdev hotkeys to the same shortcut you already use.
- ↗To Talon: move if you need a polished graphical configuration experience and voice-command scripting over file-based configuration.
- ↗To a hosted cloud dictation service: move if you want cloud-grade accuracy with no local hardware requirement and no build-from-source setup.
- ↗To a packaged distro dictation tool: move if you need a one-click precompiled binary and are willing to trade local model choice and hot-in-memory speed for it.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 5 videos for “Hyprwhspr”, and we withheld 5: 5 could not be judged, because “Hyprwhspr” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Hyprwhspr.
Official links
Tools that pair well with Hyprwhspr
Common stack mates teams adopt alongside Hyprwhspr, with the specific reason each pairing earns its keep.
Speech to Text Dictation AI
Speech to Text Dictation AI turns your iPhone or iPad into an on-device dictation and transcription tool.
Soniox
Soniox speech-to-text API delivers real-time STT, TTS, and translation in 60+ languages through one unified developer platform.
Speechmatics
Speech-to-text API for accented, multilingual audio — real-time at 350ms or batch, deployable on-prem, on-device or in the cloud
Featured Head-to-Head Comparisons
Hyprwhspr vs Soniox
Hyprwhspr is a free, private, Linux-only dictation tool for Wayland users who want local processing and GPU acceleration. Soniox is a paid cloud API for developers building real-time multilingual voice apps with low latency and enterprise compliance. Choose Hyprwhspr if you're on Linux and need hands-free typing; choose Soniox if you need a scalable, language-rich speech API.
Hyprwhspr vs Retell Ai
Hyprwhspr and Retell AI serve completely different niches: Hyprwhspr is a free, privacy-first dictation tool for Linux power users, while Retell AI is a paid enterprise platform for automating phone conversations. Choose Hyprwhspr if you need offline, system-wide speech-to-text on Wayland. Choose Retell AI if your goal is scaling call center operations with AI agents.
Hyprwhspr vs Voiceitt
Choose Hyprwhspr if you're a Linux Wayland user needing fast, private, local dictation with flexible modes and GPU acceleration—it's free and developer-friendly. Choose Voiceitt if you or your users have non-standard speech patterns (e.g., cerebral palsy, ALS, accents) and require a trained, inclusive voice AI with meeting captioning integrations. They solve completely different problems.
Alternatives to Hyprwhspr
View all
Speech to Text Dictation AI
Speech to Text Dictation AI turns your iPhone or iPad into an on-device dictation and transcription tool.
Soniox
Soniox speech-to-text API delivers real-time STT, TTS, and translation in 60+ languages through one unified developer platform.
Speechmatics
Speech-to-text API for accented, multilingual audio — real-time at 350ms or batch, deployable on-prem, on-device or in the cloud
Frequently Asked Questions
Used Hyprwhspr? Help shape our editorial sentiment research.