openwhispr

OpenWhispr/openwhispr

Privacy-first voice dictation, meeting transcription, and notes

PersonalEnterpriseNotes & docsLocal inferenceDesktopMobilePermissive

At a glance

stars on GitHub
10k
contributors
112+
last commit
latest release, 7 days ago
v1.10.4
license
MIT

About

OpenWhispr is a cross-platform voice-to-text application for dictation, meeting transcription, and note-taking. It supports fully offline speech recognition using local models like Whisper and NVIDIA Parakeet, as well as cloud AI providers. It runs as a desktop application on macOS, Windows, and Linux, with an accompanying mobile app.

From the repository: “Voice-to-text dictation app with local (Nvidia Parakeet/Whisper) and cloud models (BYOK). Privacy-first and available cross-platform.”

Latest release: 1.10.4, . Its files were downloaded 402,775 times.

All files (19)

Windows

macOS

Linux

Other files

These files come from the project's GitHub release. DigGitHub does not check them.

openwhispr and its alternatives: GitHub stars, last commit, license and worth score
ProjectStarsLast commitLicenseWorth
openwhisprThis project
10kMIT77
omiAI assistant capturing audio and screen to build personal memory
14kMIT84
karakeepSelf-hostable bookmarking app with AI tagging and search
30kAGPL-3.084
GenOfficeOpen-source AI office suite
11kApache-2.087 (best)
SiYuanPrivacy-first knowledge workspace
47k (best)AGPL-3.085

▲ marks the most stars and the best worth score.

openwhispr.com
Preview of openwhispr
Preview of openwhispr at openwhispr.com

Topics