Start with the source.
Paste one public /reel/, /reels/, /p/ or /tv/ video URL, including username-prefixed links. Or select one saved audio or video file. No private links, Stories, profiles or multi-video carousels.
FROM WATCHING TO WORKING
Connecting to transcription serviceInstagram Transcript
01 / SOURCENo account needed
1 Read video
2 Recognize speech
3 Take the text
Ready when you are.
02 / TRANSCRIPT
Your words land hereFrom a voice to a useful draft.
Add one video to get the spoken text.
No sample transcript. Just your result.
Speech recognition is a draft. Check names, numbers and quotations against the original.
Post caption · separate from spoken audio
A SMALL TOOL. A CLEAR PURPOSE.The words you need,
The words you need,
without the replay.
Use this Instagram transcript generator to turn Instagram Reels to text, or take an Instagram video to text from a saved file. Your current result stays editable until you leave the page.
Real speech recognition.
Our server downloads the public media and runs Whisper. It does not require existing captions. Transcription and English translation use the small multilingual model. Explicit Cantonese transcription uses large-v3-turbo; select Cantonese manually. Processing time depends on the clip and the server.
View model language coverage
Loading the installed model’s language list…
Model coverage is not an accuracy guarantee. Only the test languages reported with this delivery have been tested here. English translation is not verified for every source language.
Read, correct, take it with you.
Edit individual transcript segments, then copy plain text or timestamped text. TXT, SRT, VTT and Markdown downloads use your edits. Timings come from the recognition model; they are not manually verified. The post caption, if available, is shown separately.
A few useful answers.
What can this tool read?
Public Instagram video links that Instagram makes accessible to this machine, or your own saved media. A valid-looking URL can still be unavailable, restricted, removed or rate-limited. No login walls are bypassed.
Does it read text on screen or identify speakers?
No. This version transcribes spoken audio with segment timestamps. It does not do image OCR, carousel analysis, speaker identification or AI chat. The post caption is separate metadata, not recognized speech.
What happens if there is no speech?
The tool reports that no clear speech was detected. It does not substitute a sample. Music and overlapping voices can still confuse recognition, so review every transcript before using it.
Where does the video go?
Link processing requests the public media from Instagram. Uploaded files travel through Cloudflare to our transcription server. Temporary media files are deleted after processing. The server keeps only the current result in memory, clears it when another job starts, and expires it after ten minutes. Nothing is saved as transcript history.