Lipflow

Lipflow: Desktop software that transcribes silent mouth movements into text at a cursor position using webcam input and local machine processing.

Lipflow operates as a transcription tool that reads silent speech through a webcam to type words at a cursor. Users hold a designated key, mouth words silently, and the application processes the video input locally to output text. It supports setups for macOS, Windows, and Linux, utilizing local models to recognize lip movements and offer optional LLM cleanup for improved accuracy. The software improves performance by training on specific facial movements and adapting to individual phrasing patterns. Integration options include Claude, Ollama, and ChatGPT via the Codex CLI to correct ambiguity in speech recognition.

Lipflow at a glance

Best for: Software engineer, Writer
Runs on: macOS, Windows, Linux
Pricing: Free
Released: September 2026
Category: AI productivity tool
Website: github.com/amywork777/lipflow

Key features

Transcribes silent speech via webcam video
Performs processing locally on the machine
Offers whisper mode combining lip reading with audio
Trains on specific facial movements
Integrates with Claude, Ollama, and ChatGPT for correction
Allows custom word additions for name recognition
Imports phrasing history from dictation databases
Handles text entry in any application with cursor support

Use it for

Dictating text without audio input
Correcting recognized words to train the model
Transcribing pre-recorded video files
Maintaining privacy through offline local processing

Alternatives to Lipflow

talkflow — Transcribes speech entirely on your device with no word limits or cloud accounts
Whisper Dictate — Runs speech models locally on your GPU so audio never leaves your machine.
LocalBolo — Private, on-device dictation that works in any app without cloud data usage.
SpeechFast — Dictate anywhere on Mac with 100 percent local processing and no cloud servers.
Whistle Translator — Processes audio entirely offline on your computer without accounts or cloud uploads.
Voxscribe AI Voice Dictation — Transcribes speech locally on your PC for complete privacy without sending audio to a server.
WhisperX — It enables 70x faster, batched transcription with high-accuracy alignment.
Tuxwhisper — Runs locally without cloud processing and types directly into any application.
OpenLess — Keeps data local and allows the use of your own preferred ASR and LLM providers.
UnCode — Runs entirely on your machine without cloud services.
faster-whisper — Transcribes audio faster with lower memory usage than standard implementations.
Pasteface — Transcribe speech locally or via OpenAI directly from your terminal with named tabs.

All Lipflow alternatives