Convert2EPUB

I keep publisher PDFs that read fine on a laptop and fall apart on a phone because the columns are fixed. The split between a web tool and a desktop app is the part I would actually use: a 400-page scan is something I would rather OCR locally than upload. Does the desktop build run fully offline, and can it export one EPUB per detected chapter instead of a single giant file? That would make long course readers much easier to sideload onto a Kindle.

Hey people, My girlfriend had started reading on her Kindle, but converting PDFs kept giving her books that were awkward or impossible to read. I’m a developer, so she eventually told me to build something better. Posted it on Reddit forums and got 1000+ upvotes, and the feedback helped me improve it.

The watermark and running-header stripping is the feature I would want a preview for before trusting a whole course pack. A lot of lecture PDFs put the real document title in the header, and an aggressive strip can delete it from every page before the EPUB is built. Is there a before/after view that shows which lines will be dropped, and does formula recognition keep equations as reflowable text rather than a flat image so a Kindle can still resize them?

For scanned textbooks, a small sample EPUB showing how a low-confidence OCR word is handled would help readers judge the result. Are uncertain words flagged for review, or silently included in the export?

The watermark and running-header stripping is the feature I would want a preview for before trusting a whole course pack. A lot of lecture PDFs put the real document title in the header, and an aggressive strip can delete it from every page before the EPUB is built. Is there a before/after view that shows which lines will be dropped, and does formula recognition keep equations as reflowable text rather than a flat image so a Kindle can still resize them?

Converting files into readable ebooks sounds helpful. Which file formats does Convert2EPUB support, and does it preserve the original formatting, images, and table of contents?

Removing repeated headers, footers and watermarks is the detail that matters most for course PDFs, since those usually break reflowable text. Formula and code recognition is also ambitious. How does it handle scanned pages with two columns, and can you review the detected chapter breaks before export?

How does it handle two-column PDFs like research papers? That's usually where converters mix up the reading order.

Great tool! Converting PDFs and other ebook formats into clean, readable EPUB files can make reading much more comfortable across different devices. I especially like the idea of making existing files feel like proper ebooks. Simple and useful solution!

Great tool for converting messy PDFs into clean, Kindle-friendly EPUBs. Love the readable formatting and handy OCR feature!

The reflowable-vs-PDF-wrapper distinction is the whole game here. One thing I'd want to check before trusting it on a textbook: how it rebuilds the table of contents and chapter navigation — a "real EPUB" lives or dies on whether the TOC actually jumps to sections on a Kindle. Do you detect headings to generate a nav spine, or rely on the source PDF bookmarks?

Removing repeated headers, footers and watermarks is the feature I'd care about most, since those are what usually pollute the text flow in converted course PDFs. Reflowable output also depends on getting reading order right. How does it handle two-column academic papers with footnotes and figures, and can it keep footnotes linked rather than dumping them mid-paragraph?

The OCR plus removal of repeated headers/footers is what makes scanned course PDFs actually usable on a Kindle. It would help to show a before/after sample of a scanned page on the listing.

Rebuilding a reflowable EPUB rather than wrapping PDF pages is the right approach, especially for course PDFs. One useful addition would be a short before-and-after sample showing how a multi-column PDF and its footnotes come out, so readers can judge the layout before converting a long document.

Rebuilding into actual reflowable EPUBs instead of flat page wraps is the game-changer most converters miss. The header/footer stripping for course PDFs is exactly what students need. Desktop offline option is a smart touch for privacy. Curious how it handles multi-column academic papers with footnotes - that's usually where the layout breaks.

Does the desktop build run fully offline, and can it export one EPUB per detected chapter instead of a single giant file?

was looking for something like this, thanks to this dev

I like that this focuses on rebuilding documents into a proper reflowable EPUB instead of just converting the file format. The chapter structure is especially useful for longer PDFs.

The automatic removal of repeated headers, footers, and watermarks stands out, since those are what usually clutter converted PDFs on a Kindle. How does the tool handle multi-column academic papers with footnotes? A practical suggestion would be a quick preview of the reflowed text before download, so people can catch layout problems early.

This is a genuinely useful tool! Converting PDFs into clean, reflowable EPUBs can make a huge difference, especially for students and people who read on Kindle. Love the OCR and formatting cleanup features. Great work! 👏

Reflowable EPUB plus stripping repeated headers, footers, and watermarks is what actually makes a course PDF readable on a Kindle instead of a zoomed scan. Formula and code recognition is the less obvious piece: most converters flatten those into images, so keeping them as text would matter for technical papers. I'd want to see how the OCR path treats a two-column scanned journal page before relying on the web tool for a whole reading list.

Nice origin story. Question on the OCR path: does it keep footnotes linked to their markers, or do they end up inline in the text? For academic PDFs that's usually what makes or breaks reading on a Kindle.

24/7 support from docs is the real win here - no more queueing users for live chat. The challenge is usually training the model on proprietary content, but this looks like it handles that workflow cleanly. Worth testing if you have SaaS or internal tool documentation that gets repetitive questions.

This could be helpful for people converting documents into ebooks. Does Convert2EPUB support batch conversion, and can users customize the layout and formatting for different e-readers?

Use this tool regularly

Most converters just wrap PDF pages; rebuilding a reflowable EPUB is the hard part. Formula and code recognition is especially useful for technical PDFs that usually become unreadable images on a Kindle. Does the TOC come from PDF bookmarks, or do you detect headings to build navigation?

Love that this started from a real reader's frustration. PDF-on-Kindle is a problem most people with an e-reader have quietly given up on. The automatic cover and metadata detection is a nice touch too, since that's what makes a converted file look like an actual book in the library instead of document_final_v3. How does it handle two-column academic papers with footnotes: do the footnotes become linked endnotes, or stay inline where they were on the page?

Stripping repeated headers and footers is exactly where most PDF to EPUB tools fall apart for me. How does it handle two-column papers with footnotes? Does the reading order survive the conversion?

Very helpful tool, I face this problem quite often

the automatic cover and metadata detection is what caught my eye. can you fix the title and author by hand before export if it guesses wrong on a scanned pdf?

Stripping repeated headers, footers and watermarks is what makes a converted course PDF actually readable on a Kindle. How does the OCR handle two-column academic papers with footnotes, and does it keep the reading order correct? A before/after preview of a few pages before converting would help users judge quality.

very great article Posted it on Reddit forums and got 1000+ upvotes, and the feedback helped me improve it.

Tried converting a messy PDF with footnotes — the fact that it rebuilds actual ebook structure instead of a flat image scan is the killer feature. One suggestion: an option to keep original chapter breaks when the source has inconsistent headings would save a cleanup pass.

Really like the focus on rebuilding structure rather than just converting pages — stripping repeated headers/footers and detecting real chapters is what separates a readable EPUB from a mess. The formula and code recognition is a nice touch for technical PDFs. One thing I'd love to see: a side-by-side before/after preview for scanned documents so users can check OCR quality before downloading. Congrats on the launch!

A thoughtful way of getting people to read more. Nowadays everyone's on different devices throughout the day and most apps aren't optimised for long reading sessions. Really helpful!

Fixed-column PDFs on a phone are a real pain, so this solves an everyday problem. Love that there's a desktop app for people who don't want to upload private documents. How well does it handle tables and footnotes? That's usually where PDF-to-EPUB converters break the layout.

Converting PDFs to reflowable EPUBs usually destroys technical formatting, so having dedicated formula and code recognition listed in the features is huge! How does the OCR model handle inline math equations versus multi-line blocks—does it preserve them as MathML or render them as images?

Solid little tool if you're tired of ugly PDFs on your Kindle. Clean EPUBs, OCR that actually helps, and formatting that feels readable.

As someone building a study app, the formula and code recognition is the part I'd look at first. When an equation can't be rebuilt cleanly, does it fall back to an image of that region, or to text? For textbooks a half-converted equation is worse than a picture of it. Also curious how you tell a repeated page header apart from a running chapter title that a reader might actually want to keep.

How do you decide what counts as a repeated header? Is it by its position on the page or by text repeating across pages? The chapter title in the header changes every few pages, so I’m wondering how you keep it from being cut as noise.