Video Transcript Generator
LOCAL WHISPER + SUBTITLE CLEANUP

Video Transcript Generator

Transcribe short English recordings with Whisper in your browser, or turn SRT, VTT and pasted captions into clean, downloadable text.

✦ Free to use↗ Export your result○ No account required
01

Make it yours

Up to 2 MB. If selected, this file takes precedence over the pasted text.
English speech; up to 60 MB and 10 minutes. WAV/MP3 work most reliably. No file upload.
Only overlapping timed captions are compared. Ordinary repeated speech is preserved.

Choose your file, then select an action. See the method below for processing details.

02

Your result

Ready when you are

A little input.
A useful starting point.

Paste a WEBVTT file exported from your editor, choose Continuous paragraph, and prepare the transcript. For original audio, switch to Whisper, select a short MP3 and explicitly start the model download.

Example · customize the fields to begin
UNDER THE HOOD

Imports use a local SRT/VTT parser with explicit timing validation. Audio is decoded and resampled to 16 kHz mono, then transcribed in a worker using Transformers.js 3.7.3 and Xenova/whisper-tiny.en at a pinned model revision. The model is based on OpenAI Whisper; this site is independent. Review machine output for errors.

A BETTER STARTING POINT

Get more from your result.

Read the full guide ↗
01

Choose the fastest source you already have

If your video editor or caption provider has already produced SRT or VTT, import that file instead of transcribing the audio again. The cleaner removes cue numbers, display settings and caption markup while…

02

Run Whisper locally with realistic expectations

Whisper tiny.en is an English speech-recognition model. The first transcription downloads model files and a runtime, then processes decoded audio on your device using WebAssembly. Short, clear clips are the…

03

Review words before reusing them

A transcript is a draft, especially when a recording contains music, accents, overlapping voices or silence. Check names, technical terms and numbers against the original media. Whisper can generate plausible…

GOOD TO KNOW

A few useful answers.

Is my video uploaded to a server?

No. This interface decodes and transcribes the selected file locally. Model and runtime downloads contact Hugging Face and jsDelivr, but the audio is not attached to those requests.

Will it work offline?

Subtitle cleanup works after the site has loaded. Whisper requires internet on the first run and may need it again if the browser evicts cached files. Offline model availability is not guaranteed.

Can I paste a YouTube link?

No. Supply a local audio/video file you can use, or export subtitles from an authorized source. This tool does not download videos or bypass platform restrictions.

Which languages and speakers are supported?

This version uses the English-only tiny.en model. It does not identify speakers, translate speech or guarantee word-level timing.

Why can’t my video be opened?

Audio decoding depends on browser and codec support, not just the filename. Try exporting the audio as WAV or MP3. Files are also limited to 60 MB and ten minutes.

Can a plain transcript become subtitles?

Not accurately without timing. SRT/VTT export appears only for imported timed cues or Whisper output with estimated timestamps.