OpenWhispr /openwhispr
Voice-to-text dictation app with local (Nvidia Parakeet/Whisper) and cloud models (BYOK). Privacy-first and available cross-platform.
★ 11,620⑂ 1,534👁 25⑃ 1,996● 1,159▤ 102
+60
Daily
+990
Weekly
+2,311
Monthly
11,620
stars
1,534
forks
25
watchers
1,996
prs
1,159
issues
102
releases
README
OpenWhispr The open-source and free alternative to WisprFlow and Granola. Privacy-first voice-to-text dictation with AI agents, meeting transcription, and notes. Cross-platform for macOS, Windows, and Linux. Website · Docs · Download · API · Changelog --- OpenWhispr turns your voice into text, notes, and actions from your desktop. Press a hotkey, speak, and your words appear at your cursor. Choose between fully private offline transcription with local speech-to-text models like Orukeet, Whisper, NVIDIA Parakeet, and Cohere Transcribe — where your audio never leaves your device — or cloud processing for speed. No data collection, no telemetry, fully open source. Download | Platform | Download | | --------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | macOS (Apple Silicon) | .dmg | | macOS (Intel) \ | .dmg | | Windows | .exe | | Linux | .AppImage / .deb / .rpm / .tar.gz | \ On Intel Macs, live speaker identification and voice fingerprinting are unavailable: they depend on ONNX Runtime, which stopped shipping macOS x86_64 binaries in 1.24. Meetings still record and transcribe normally, and notes search falls back to keyword matching instead of semantic search. On macOS or x86_64 Linux, you can also install the community-maintained Homebrew cask: bash brew install --cask openwhispr Features - Voice dictation — global hotkey to dictate into any app with automatic pasting - Dictation translation — dedicated hotkey to dictate in one language and paste the text in another - AI agent — talk to GPT-5, Claude, Gemini, Groq, Tinfoil, OpenRouter, or local models with a named voice assistant - Voice Assistant hotkey — dedicated hotkey that sends what you say straight to your AI assistan
Top issues
#462Whisper hallucinations ("Thank you for watching!") during silence — needs smarter VAD for cloud backends14 comments · 2026-03-18#459Android Support12 comments · 2026-03-17#2384Auto paste stopped working10 comments · 2026-09-28#1860Support both Tap and Hold Activation Modes simultaneously8 comments · 2026-08-26#1400Add a way to cancel dictation in hold/push-to-talk mode8 comments · 2026-07-30#1280[BUG] Note Recording ignores the custom Cloud Providers endpoint — always connects to hardcoded wss://api.openai.com/v1/realtime7 comments · 2026-07-22#2640KDE Wayland: voice-agent popup accepts no clicks or keyboard and cannot be closed6 comments · 2026-10-13#2013BUG Start OpenWhispr6 comments · 2026-09-03
Discussions
#2854Can numbers be transcribed as a number, not spelled out1 comments · 2026-10-21#2596Adjustable context length for local models1 comments · 2026-10-09#2591Why are two-modifier shortucts not supported on Mac?0 comments · 2026-10-09#2310Quick language switching option to select the language.0 comments · 2026-09-23#2230Lite Version + Disable Menu Bar Icon Option0 comments · 2026-09-17#2174We are hiring!0 comments · 2026-09-13#2170Support uploading video files for transcriptions0 comments · 2026-09-13#2039Make auto pasting/linux-fast-paste configurable0 comments · 2026-09-05
Latest release
1.11.0 v1.11.0 2026-10-29OpenWhispr 1.11.0 Teams can now share a dictionary, the assistant can follow skills you keep for recurring work, and a bunch of new models. 🌟 Highlights - 📚 Shared team dictionary — Business and Enterprise members get their team's words in dictation and meeting notes. Your own words still come first. - 🧩 Skills — Save instructions for recurring work in Integrations → Skills. The assistant picks a fitting skill on its own. - 🤖 New models — GPT-6.1 Sol, Gemini 3.8 Flash, Claude 5.5 on Bedrock, Gemma 4, Qwen3.8 and more. Older models keep working if you already use them. - 🏢 Workspac