Whisper Local
Free, open-source AI dictation for Windows and macOS. Runs Whisper locally, keeps audio on your machine, and supports voice editing and Ollama rewrites.
Whisper Local is a free, MIT-licensed dictation app for Windows and macOS, aimed at people who want voice typing without cloud processing. It's an alternative to Wispr Flow and Dragon that inserts speech into the active text field, including in chat apps, browsers and code editors. It doesn't require an account or API key.
Transcription runs on your computer using OpenAI Whisper through faster-whisper and CTranslate2. It works on CPU, with GPU acceleration through NVIDIA CUDA or AMD ROCm. After the initial model download, dictation works fully offline. Audio stays on your machine, and the app sends no telemetry; optional update checks use an internet connection.
A global push-to-talk hotkey controls recording, with hands-free and continuous modes for longer notes. Spoken punctuation, corrections and reusable text snippets reduce the editing needed after dictation. Per-app rules can give email full-sentence formatting while keeping code and terminal input verbatim. Voice commands can trigger keyboard shortcuts or run shell commands, and translation mode converts speech into English.
Optional Ollama integration uses a local LLM to clean up transcripts or rewrite selected text according to a spoken instruction. Basic formatting and voice editing work without an LLM. A searchable transcript history lets you recover earlier dictation and turn recurring recognition errors into saved vocabulary corrections. The local OpenAI-compatible transcription API also supplies speech-to-text to Cursor, VS Code Continue, Open WebUI and n8n.
