OpenVoiceFlow
Push-to-talk dictation that transcribes on device with Whisper and pastes the text into whichever app you are using.
AGPL-3.0+ · Downloads from openvoiceflow.com
- License
- AGPL-3.0-or-later
- Open source
- Source
- GitHub
- shimoverse
- Verified
- 2026
About OpenVoiceFlow
OpenVoiceFlow lives in the menu bar: hold a key, speak, and release to have the transcript pasted at the cursor. Transcription runs locally through WhisperKit with a choice of Whisper model sizes, and words appear in a small floating display while you talk. A personal dictionary, spoken snippets and per-app writing styles adjust the result.
An optional cleanup step, off by default, can polish the text through OpenRouter, OpenAI, Anthropic, Groq or a local Ollama model. Anonymous aggregate usage sharing is on by default and can be switched off. The AGPL code carries Section 7 attribution and naming terms for forks, recorded in its NOTICE file. Signed builds for macOS 14 and later come from the website and GitHub releases.
Information
- Source code
- shimoverse/openvoiceflow
- License
- AGPL-3.0-or-later
- Website
- openvoiceflow.com
- Download
- openvoiceflow.com
- Category
- AI
- Evidence checked
- Tags
Maco links to each project’s official download. Check its release notes for macOS version and Apple silicon support. The verification date records our license and source review, not a security audit.
Sources and image credits
- github.com/shimoverse/openvoiceflow
- github.com/shimoverse/openvoiceflow/blob/main/LICENSE
- github.com/shimoverse/openvoiceflow/blob/main/NOTICE
- github.com/shimoverse/openvoiceflow/releases
- openvoiceflow.com/download.html
- icon.png: Shimoverse Studios. AGPL-3.0 repository asset; the license grants no rights in the OpenVoiceFlow name or logo.
License and artwork notices