HyperVoice is a dictation hotkey for Windows, macOS, and Linux. It is built for turning speech into text at the cursor in any app, using a speech engine that runs locally so audio never leaves the machine.
Its core loop is simple: press the hotkey, speak, and let the text land wherever focus is active. The app can be triggered with Ctrl+Shift+Space from any application, and the key can be rebound. It supports tap-to-toggle or hold-to-push-to-talk behavior, runs from the system tray, and shows a floating always-on-top pill with status changes while it listens, transcribes, and inserts text. HyperVoice uses Whisper or Parakeet on the user’s machine, supports transcription without internet access, and uses GPU acceleration through Vulkan on Windows and Linux for sub-second results, with CPU fallback when no GPU is available. It also keeps a searchable local history, shows usage statistics, and includes a custom dictionary for jargon and names.
The product can paste dictated text into any application with a text field, including editors, email, chat, terminals, and other system-wide text inputs. Optional AI cleanup is available through cloud processing: there are seven built-in rewriting styles, support for custom prompts, and cloud use through the user’s own OpenAI or Anthropic API keys or through HyperVoice Cloud on Pro. That cleanup is described as optional and off by default.
HyperVoice is offered as a free download with 500 words per day and no credit card required. The page also describes a lifetime license with no subscription, and notes that the Mac release is Apple-signed and notarized, runs on Apple Silicon, and opens in one click. It is presented as a cross-platform dictation application with on-device speech transcription.
HyperVoice is a Speech to text project. It focuses on transcribing speech to text locally without sending audio to the cloud. It is built as a consumer product for professionals and individuals needing dictation. HyperVoice follows a freemium model, with paid plans starting at $7. It ships for the command line, Windows, Linux, and macOS.
HyperVoice first shipped in 2026. Among its 5 catalogued features are voice dictation, hotkey activation, and local speech-to-text.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do