Download with confidence
Every release publishes installers, signatures, updater manifests, and checksums from one stable location.
VILab is a desktop-first voice input platform with local speech capture, cloud text transforms, self-hosted HTTP APIs, and a release pipeline built for real software distribution.
This repository is the public face of VILab. The private source repository stays private, while this site handles the parts users need: installers, release notes, public documentation, and the updater feed.
Every release publishes installers, signatures, updater manifests, and checksums from one stable location.
The public site reuses the same docs build that the desktop app serves locally under /docs/.
Development stays in the private repo. Public users only see the polished distribution surface and changelog.
VILab is more than dictation. It is a reusable voice layer that can power local desktop workflows, LAN services, and future multi-client integrations without forcing one model provider or one deployment shape.
Start with a local host app that handles shortcuts, recording, transcription, cleanup, and operator feedback.
Expose the same core through HTTP endpoints so scripts, LAN clients, and internal tools can reuse it.
Versioned notes, stable update feeds, and public installers make the platform easier to trust and easier to support.
The latest stable versions stay documented here so users can decide when to upgrade and what changed.
Moved desktop release distribution to the public repository so installers, release notes, and updater assets can be accessed without private GitHub access.
Read release notesReworked macOS global shortcuts with an event-listener backend so dictation can trigger more reliably on Apple Silicon Macs.
Read release notes`VILab 0.1.5` introduces traceable transcript output variants and a browser-hosted Prompt Lab for internal prompt and model evaluation.
Read release notes