Reads Pages, Does Not Download Them
AITuber extracts the writing on a page and creates a new video about it. It is a content tool, not a media downloader, and it never copies an existing video file.
Paste a public link. AI reads the page, writes a script from what it says, and builds a narrated video with visuals and captions.
Sample video. Your result will vary based on the style, voice, and settings you choose.
No editing skills. No complex software. Just describe what you want.
Any public page made mostly of text: an article, a guide, documentation, or an announcement.
Summarize for long pages, Read it out for short ones. Set the duration and add a note if you want the script steered.
The page is fetched and read, then a script, visuals, voiceover, and captions are generated. Export the MP4 or publish to YouTube.
Professional tools, zero learning curve.
AITuber extracts the writing on a page and creates a new video about it. It is a content tool, not a media downloader, and it never copies an existing video file.
Navigation, ads, cookie notices, newsletter popups, and related-article rails are removed before anything is written, so the script comes from the actual article.
Many pages refuse plain automated requests. AITuber fetches through infrastructure built for it, so ordinary publisher and documentation sites read correctly.
Condense a 4,000 word guide into 60 seconds, or narrate a short post close to as written. One toggle, and it changes what the script does with the source.
No waiting on a preview before you can continue. Paste, set the length, and submit. Fetching, reading, and writing all happen in the background.
Read an English page and narrate in Spanish, Hindi, or Japanese. The source language and the output language do not have to match.
Captions track the narration word by word, which is what makes the video work on feeds that autoplay muted.
Download up to 4K with no watermark, or publish straight to a connected YouTube channel.
First, the thing most people arrive here looking for: this is not a video downloader. It does not pull an existing video off a link and save it as a file. AITuber does the opposite. You give it the address of a page made of words, it reads that page, and it produces a brand new video about what the page says.
Here is how it runs. Paste a public link and AITuber fetches the page, strips the navigation, ads, cookie banners, and related-article rails, and keeps the body text. It then writes a script from that text, generates a visual for every scene, narrates it with one of 1,300+ voices, and burns in word-synced captions. You choose the length first, so the script is written to fit rather than cut off at the end.
Two modes decide how faithful it stays. Summarize it reads the whole page and writes a tighter script around the main points, which is what you want for anything long. Read it out narrates the text close to as written and trims it to your chosen duration, which suits short pieces you want spoken accurately.
There are limits worth knowing before you paste. Pages behind a paywall or a login cannot be read, because the fetch sees what a logged-out visitor sees. Pages that are mostly video, images, or an app shell have little text to work from and produce weak results. A normal article, guide, or documentation page is the sweet spot.
Open the link in a private window. If you cannot read the article there, neither can we, because the fetch has no account and no cookies.
A homepage is a list of headlines with no substance. Link directly to the piece you want covered and the script has something to work from.
Read it out on a long page means the narration is trimmed at your duration and the ending never arrives. Summarize reads it all and decides what earns the time.
A note like "ignore the pricing section" or "focus on the second half" is the fastest way to change the output, and it takes one sentence.
If the page is not yours, name it under the video. It is both the fair thing to do and it consistently performs better in comments than an uncredited summary.
No. It does not download or save video from a link. AITuber reads the text on a web page and generates a new video about that content, with its own script, visuals, and narration.
Public pages made mostly of text: articles, blog posts, guides, documentation, announcements, and reference pages. Pages that are mostly video, images, or an app interface have too little text to work from.
No. The fetch sees exactly what a logged-out visitor sees, so a paywalled article returns only the preview above the wall. If the page is gated, copy the text you have rights to and use the script input instead.
Same answer. Anything that needs an account, a session, or a private link cannot be read. Internal wikis, dashboards, and members-only pages are all out of reach.
You are responsible for what you publish. Summarising a public article and crediting the source is normal practice, but reproducing a full copyrighted piece word for word is not. Use Summarize rather than Read it out for other people's work, and name the source.
Visuals are generated scene by scene from the script, not scraped from the page. Nothing is lifted from the site you linked, so there are no image rights to worry about.
Yes. Set the target duration before generating, from 15 seconds up to 20 minutes. The script is written to that length rather than cut down to it afterwards.
The result will be thin, because there is nothing to work from. Landing pages, photo galleries, and video pages are poor sources. Aim for at least a few hundred words of real writing.
Yes. The page can be in one language and the narration in another, across 50+ languages and 1,300+ voices. This is a common way to localise documentation.
No. The 100 credits you get at signup cover several videos from links, and no card is requested to claim them. A plan only becomes necessary once you are generating regularly.
Create videos for other popular niches
Join 75,059+ creators using AITuber to make professional url to video videos with AI.
No credit card required