OpenWhispr

OpenWhispr is an open-source desktop dictation tool built by Gizmo Labs Inc. that turns voice into text, notes, and actions across any application. It supports fully offline transcription using local models (OpenAI Whisper and NVIDIA Parakeet) as well as cloud-based processing, and runs on macOS, Windows, and Linux. The GitHub repository has accumulated over 3,700 stars since the project was created in June 2025.

What It Is

OpenWhispr is a cross-platform voice-to-text assistant in the category of AI dictation and meeting transcription tools. Its core job is to let users press a global hotkey, speak, and have their words appear at the cursor in any text field — in any app. Beyond basic dictation, it includes an AI Notepad for meeting notes, an AI Chat interface that can reference past meetings, and an audio upload feature for transcribing recorded files locally. The project positions itself as an open-source alternative to tools like WisprFlow and Granola.

Privacy and Local-First Architecture

Privacy is a central design principle. The project's homepage states that audio passed through cloud processing is never stored, that transcription history lives in a local SQLite database on the user's machine, and that local Whisper and Parakeet models run entirely on-device with no internet required. The codebase is publicly available under the MIT license on GitHub, which the project cites as a way for users to verify these claims independently. The tech stack includes React 19, TypeScript, Tailwind CSS v4, Electron 41, better-sqlite3, whisper.cpp, and sherpa-onnx.

Core Features

• Global hotkey dictation — speak into any app; text is automatically pasted at the cursor
• Local model support — choose from Whisper Tiny (75 MB) through Whisper Turbo (1.6 GB) or NVIDIA Parakeet, all running on-device
• Cloud transcription — bring your own API keys (OpenAI, etc.) for faster cloud-based processing
• Meeting transcription — auto-detects Zoom, Teams, and FaceTime calls with live speaker diarization and voice fingerprinting; integrates with Google Calendar
• AI Notepad — generates structured meeting notes with decisions, action items, and open questions
• AI Chat — a chat interface that can query across your meeting history
• Audio Upload — transcribe any audio file locally
• Custom dictionary — auto-learns names, jargon, and technical terms from corrections
• 100+ language support — with automatic language detection
• Public API and MCP server — manage notes and transcriptions programmatically or connect AI assistants via the MCP integration

Platform and Integration Reach

OpenWhispr works system-wide on macOS (Apple Silicon and Intel), Windows, and Linux (AppImage, .deb, .rpm). Because it pastes text at the cursor, it works in any application that accepts text input. The homepage lists compatible apps including ChatGPT, Claude, Cursor, Google Docs, Gmail, Slack, Microsoft Teams, Superhuman, Grammarly, iMessage, and X. Use-case guides cover developers, writers, healthcare and legal professionals, accessibility users, offline workflows, and translators.

Update: v1.7.2

The latest release is v1.7.2, published on May 20, 2026. The repository was created in June 2025 and has been actively maintained, with the last push recorded in June 2026. The project is written primarily in TypeScript and has 518 forks and 205 open issues as of the latest data. The roadmap visible on the pricing page includes an iOS mobile app listed as "Coming Soon" on the Pro tier.