Free Speech to Text

About Free Speech to Text

Free Speech to Text converts audio, video, and live microphone input into text with word-level timestamps. Everything runs locally in your web browser using WebGPU (with WebAssembly fallback).

Open source base

This app is derived from the whisper-word-timestamps example in Hugging Face transformers.js-examples (Apache-2.0). The project LICENSE file is retained in the repository.

Model

We use onnx-community/whisper-base_timestamped, a community ONNX conversion of OpenAI's multilingual whisper-base. Whisper weights are published by OpenAI under MIT on GitHub. ONNX weights are loaded from Hugging Face at runtime and cached in your browser — they are not vendored in this repository.

Credits

Explore more free tools at web-free.site. Pair transcription with Free Text to Speech for a full read-aloud workflow.

← Back to transcribe