Instagram Transcript Generator

What it supports

A video or audio file from your device: MP4, MOV, M4A, MP3 and WAV. Speech is transcribed by Whisper running inside this page. The model ships with 99 language codes; pick the one being spoken and it stops guessing.

How to use it

1. Save the clip to your device. On Instagram that is the … menu on the post, then Save or Share → Save video.
2. Choose the file above. Transcription starts on its own — there is no second button to press.
3. Read it, copy the text, or download .srt / .vtt for your editor.

Limits, stated plainly

• The file has to be on your device. This page is served as plain files and cannot read instagram.com from another origin, so there is no link box — download the clip first, then upload it.
• Private accounts are not a problem here: whatever you can save, you can transcribe.
• The model is small. It is tuned for clear speech and finishes a 60-second clip in well under a minute on a laptop. Heavy music or overlapping voices lower accuracy; no accuracy percentage is printed here because none was measured.
• Language coverage is 99 language codes — counted from the model's own lang_to_id table. No larger number is claimed, and the tool does not guess the language: you pick it.
• The first file you pick downloads 38 MB. The speech model is fetched once, then cached by your browser. Nothing is sent anywhere: the audio is decoded and transcribed locally.
• No history is stored. There is no server to store it on.

Questions

Do I need an account?

No. There is no sign-up, no login and no pop-up.

Where does my audio go?

Nowhere. It is decoded and transcribed in this tab; closing the tab discards it.

Can I get timestamps?

Yes — switch to SRT or VTT and every line carries its start and end time.

Can it read a clip that is only music?

It returns whatever speech it finds. Pure music yields little or no text.