Openwhispr

OpenWhispr is open source dictation and meeting notes software for macOS, Windows, Linux and iOS. It turns speech into text in any application, records meetings without a bot joining the call, and cleans up or drafts text from spoken instructions such as "clean this up" or "draft an email to Mike". Speech to text can run locally on the user's own computer using Whisper models in sizes from Tiny to Turbo, or through cloud transcription, or with the user's own API keys. The vendor states that local dictation sends nothing to it or anyone else, that cloud recordings are deleted once transcription finishes, and that transcriptions are not used to train models. More than 100 languages are supported, along with a custom dictionary, audio upload, AI chat, and integrations. The code is published on GitHub under the MIT license, and the site lists HIPAA, ISO 27001, SOC 2 Type II and GDPR. It suits individuals who dictate often, professionals who need private transcription, and teams that want meeting notes without a recording bot. The Free plan costs $0 and includes unlimited local models, unlimited cloud transcription with the user's own API keys, 2,000 words per week of OpenWhispr Cloud transcription and 5 hours of meeting recordings per month. Pro adds unlimited cloud transcription, agent mode, device sync, MCP integration and 20 hours of recordings per month, at $8 per month or $80 per year. Business adds unlimited recordings, chat over your data and priority support at $16 per user per month or $160 per user per year. Enterprise adds SSO, SAML, SCIM and audit logs at custom pricing.