OpenWhispr

OpenWhispr is a dictation and meeting transcription app for people who want to write by voice and keep searchable records of conversations. It lets you choose local processing or cloud providers.

• Dictation and translation: Write into apps such as Slack, Google Docs and Cursor, or speak in one language and insert text in another; language detection handles switches, and a custom dictionary learns terms from corrections.
• Voice assistant: Ask AI to draft text or edit a selection, using local models or providers such as OpenAI, Claude and Gemini.
• Meeting transcripts: On desktop, detect Zoom, Teams and FaceTime calls, label speakers on-device, and recognize voices across meetings, with Google, Microsoft and Apple calendar integrations.
• Notes and recordings: Organize notes in folders, search by meaning, apply AI actions, and transcribe imported audio or video files.
• Sharing: Cloud sync, web sharing and team spaces let signed-in users share notes with link, domain or invitation-based access.

The app is free and open source under the MIT license, with macOS, Windows, Linux and iOS versions. Local transcription works offline and keeps audio on your device. Optional cloud transcription sends audio to an external service; you can bring your own API keys. Sync and web sharing use cloud services.

Desktop speech models include Whisper, NVIDIA Parakeet, Orukeet and Cohere Transcribe. AI reasoning, speaker labeling and semantic search can also run locally. Whisper supports Apple Metal, NVIDIA CUDA and Vulkan acceleration for AMD and Intel GPUs. Intel Macs can record and transcribe meetings, but don't support live speaker identification or voice fingerprinting; note search uses keyword matching.