AI quiz generator from video,
without uploading the video.
Bring the words that were spoken. The questions come back tied to them.
No account, no file upload, and the model never sees a single frame.
Open any question to see the line it came from. Paste 40 characters or more to start.
Video is not only YouTube.
Most of the material people genuinely need to revise never gets published anywhere. It is a seminar sitting in Panopto, a Teams call whose transcript is already in the chat, a Loom somebody sent on Friday, a conference talk with an .srt file next to the download. Every one of those systems will hand you the text if you ask, and the text is all this needs.
Zoom and Teams put a transcript tab on the recording page. Panopto and Echo360 show captions in a side panel you can select from. An .srt or .vtt file opens in any text editor. Loom prints the transcript under the player. Whichever one you have, copy it and paste it above.
What the model does, and the one thing it cannot do.
It reads the words and writes multiple choice questions from them, attaching the sentence each question came from. What it cannot do is see. There is no vision model here. No frames, no slides, no whiteboard, no diagram, no equation that was written on screen and never spoken. If the value of your recording is visual, an AI quiz generator working from video transcripts will miss most of it, and it is better to know that now than at question four.
The same limitation is why nothing gets uploaded. No media file leaves your machine, there is no processing queue, and no account accumulates a library of your recordings on our side. You paste text, questions stream back, you answer them here, and you export a file if you want to keep them.
Works from
- Zoom, Teams, Meet and Loom transcripts
- Panopto, Echo360 and other lecture capture captions
- .srt and .vtt subtitle files, pasted as text
- YouTube transcripts, copied from the panel
- Any text at all, up to 10,000 characters
Cannot
- Accept a video file or a URL
- See slides, diagrams, code on screen or handwriting
- Tell you who was speaking, or hear tone
- Transcribe audio. Bring text that already exists
- Take more than 10,000 characters in one pass
Against the tools ranking here
ScreenPal, Kapwing and Questgen all work on the video itself, which means an upload, a wait and usually an account. That is a genuinely different product, and for a teacher building a graded quiz around a video they own, it is probably the right one. MagicForm and RemNote sit closer to us but keep the result inside their own app.
This page is for the narrower case: you already have the transcript, the next lecture starts in half an hour, and you want questions out of it now.
What it can and cannot see
Can I upload an mp4?
No. There is no file upload anywhere in this tool. Paste the transcript text and it works from that, which is also why there is no queue to wait in.
Where is the transcript of a Zoom or Teams recording?
Both attach one to the recording when transcription was switched on. Open the recording page, find the transcript tab, copy the text. If nobody turned transcription on, there is nothing to copy and this cannot help you.
Will it quiz me on what was on the slides?
Only if the speaker said it out loud. The model reads text, not pixels. Anything that existed only on screen is invisible to it.
How do I know the questions match the recording?
Each one shows the line of transcript it came from, highlighted in the text you pasted. Anything the model cannot tie back to your text is dropped on the server before it reaches you.
Is it free?
Two generations a day with no account, exports included. Pro is not open for sale yet, so nothing here is a trial that expires.
Related pages
Or start from the QuizPaste home page, where the same tool takes any text you paste.
QuizPaste is an independent tool and is not affiliated with YouTube, Zoom, Microsoft, Panopto or any other platform named here. Those names describe where you can find a transcript, not a partnership.