OpenWhispr
OpenWhispr is a privacy-first voice assistant for dictation and meeting notes that transcribes speech at roughly 150 words per minute—about three times faster than typing. The desktop app is open source under the MIT license and offers three processing modes: fully local on-device transcription, bring-your-own-API-key, and OpenWhispr Cloud with immediate audio deletion. It supports 100+ languages, includes AI cleanup features, and offers meeting recording with speaker labels.
- Artificial Intelligence
- Enterprise Software
- AI Agents
- Software Only
Funding
Funding not disclosed
Founders
Founder details are not available yet.
Product
Problem
Traditional typing limits productivity, with average writing speeds around 40 words per minute, and many professionals lose context and time when manually taking notes during meetings. Existing voice dictation tools typically rely on cloud processing that raises privacy concerns about sensitive voice and meeting data being stored or used for training.
Solution
OpenWhispr is a privacy-first voice assistant available as a desktop and mobile app that transcribes speech-to-text at roughly 150 words per minute, three times faster than typing. The application offers three processing modes: fully local on-device transcription with nothing leaving the computer, bring-your-own-API-key for direct connection to third-party providers, and OpenWhispr Cloud where audio is deleted immediately after transcription finishes. An AI cleanup feature works across 100+ languages to polish raw dictation, and users can give voice commands like "clean this up" or "draft an email to Mike" to trigger actions. The desktop app is open source under the MIT license, allowing full inspection of the code and privacy claims on GitHub. Meeting recording capabilities include speaker labels, AI-generated notes, and optional sync across devices.
Target Audience
Primary users are individual professionals, developers, writers, and knowledge workers who want faster dictation without compromising privacy, plus IT teams and businesses that require verifiable open-source software and enterprise-grade security controls. The platform also serves teams needing meeting transcription and collaboration tools with admin capabilities.
Features
- On-device local dictation using open-source models (Whisper variants from Tiny at 75 MB to Turbo at 1.6 GB) with zero data transmission
- Three transcription modes: local processing, bring-your-own-API-key, and OpenWhispr Cloud with immediate audio deletion
- AI cleanup and instruction-following across 100+ languages, including commands like "clean this up" or "draft an email"
- Meeting recording with speaker labels, with 5 hours free per month and up to unlimited on paid plans
- MIT-licensed open-source desktop app with full source code available on GitHub for privacy verification
- Voice-controlled agent mode for drafting emails and chat-over-your-data on Business plans
- Custom dictionary support and offline mode for use without internet connectivity
- MCP integration and command-line interface for developer workflows and API access on paid plans