100% local open source AI speech-to-text model
Turn your voice into formatted text. It's open source and runs 100% locally. 3-5x faster than typing, and especially helpful when prompting LLMs, writing emails and sending texts. Pick your preferred model, and even edit the system prompt. Take full control.
OpenWhispr enables dictation across all apps (ChatGPT, Slack, Docs, etc.) with local AI models that never send audio to servers, achieving ~150 WPM vs ~40 WPM typing. Users retain full data privacy with zero data retention on cloud processing.
Supports 100+ languages with auto-detection and mid-conversation switching. Custom dictionary feature auto-learns specialized terms (medical, legal, technical jargon) from user corrections, improving accuracy for domain-specific vocabulary.
Open source codebase on GitHub allows users to audit code, run models locally, or bring their own API keys. All transcription history stored locally in user's database; no model training on user data without explicit consent.
OpenWhispr runs Whisper and Parakeet models locally on user hardware with no internet required. Users can choose from 5 model sizes (75 MB to 1.6 GB) based on speed/accuracy tradeoff, enabling transcription anywhere without WiFi.
AI Notepad feature automatically transcribes meetings, extracts decisions and action items, and generates structured notes. Free tier includes 5 hours/month; Pro tier adds 20 hours/month; Business tier offers unlimited meeting recordings with agent mode for deeper analysis.