Turn any YouTube video into clean Markdown.
Paste a link and get a summary, key takeaways and a timestamped transcript — even for videos without captions. No copy-paste, no browser extensions.
Your files are never stored: documents convert in your browser where possible, and anything sent to our AI travels encrypted (HTTPS) and is discarded after processing. Privacy Policy
Most YouTube-to-text tools fail the moment a video has no subtitles. CleanMD listens to the audio itself: paste a public video URL and the AI produces a structured Markdown document — a short summary, the key takeaways as bullet points, and a full transcript with [mm:ss] timestamps, split into chapters for longer videos and with speakers labeled when they're distinguishable. Perfect for turning lectures, podcasts, interviews and tutorials into searchable notes for Obsidian, Notion or your knowledge base.
Working with related formats? See also Video to Markdown, URL to Markdown, Audio to Markdown.
What you get from a YouTube link
Most YouTube-to-text tools download the captions and paste them together. That works for videos with good captions, and fails for the rest: auto-generated captions without punctuation, videos with captions turned off, or talks where the slides matter as much as the voice. The result is usually a single paragraph that runs for pages.
CleanMD gives the link to an AI model that listens to the audio itself, so captions aren't needed at all. What comes back is a document: a short summary at the top, the key takeaways, and then the full transcript cleaned of filler words, with every paragraph starting at its [mm:ss] timestamp. Longer videos are split into chapters that become headings, so a one-hour lecture turns into something you can scan and jump through.
Speakers are labelled when their voices are clearly different, as Speaker 1 and Speaker 2 or by name when someone introduces themselves. In a busy panel with people talking over each other the labels get less reliable, and we'd rather tell you that than pretend otherwise.
The summary and the takeaways are written in the language of the video, so an Italian talk gives you Italian notes. If the model's answer is cut off because it hit its length limit, which happens with long videos, the request is retried automatically instead of handing you half a transcript.
The video has to be public, and the limit is around 45 minutes. Private, unlisted and age-restricted videos can't be processed. YouTube conversion is part of Premium, because the model has to process every minute of the video and that has a real cost. For a recording you have on disk, the audio and video converters accept files directly.
How it works
- 1Paste the YouTube URL in the converter above and press Enter.
- 2The AI watches the video, transcribes the speech and structures it — summary, takeaways, timestamped transcript.
- 3Preview the Markdown and download the .md file.
Before and after
[45-minute product strategy talk] No captions available. Auto-generated subtitles: disabled. Speakers: 2 Slides shown on screen
## Summary A 45-minute talk on product strategy covering positioning and pricing. ## Key takeaways - Position against the category leader, not the whole market - Price on value delivered, not cost ## Transcript ### Positioning **[00:32]** **Speaker 1:** Let's start with why positioning fails...
Frequently asked questions
Does it work with videos that have no captions?
Yes — that's the point. The AI listens to the audio directly instead of scraping subtitles, so videos without captions (or with auto-captions disabled) convert just as well.
Which videos are supported?
Public YouTube videos up to roughly 45 minutes. Private, unlisted or age/region-restricted videos can't be processed.
Do I get timestamps and speakers?
Yes. Each transcript paragraph starts with its [mm:ss] timestamp, longer videos are split into chapters, and speakers are labeled when they're clearly distinguishable.
Is YouTube to Markdown free?
No. It's a Premium feature, because the AI has to process the whole video and that has a real cost per minute. Premium includes up to 10 YouTube, audio or video conversions and 120 minutes of media a day. Converting documents and images stays free.
Is my video uploaded or stored anywhere?
No file ever passes through our servers — we send only the public video URL to the AI, which fetches and processes it directly. Nothing is downloaded or stored after the conversion.