Here's the thing most carousel tutorials bury: a LinkedIn carousel isn't a special format. It's a document post, and a document post is a PDF — every page becomes one swipeable slide. Understand that, and the whole workflow simplifies to "make a good PDF."
The specs that matter
- Format: PDF (PowerPoint and Word uploads get converted to PDF by LinkedIn anyway)
- Slides: one PDF page = one slide; LinkedIn supports up to 300 pages, but engagement data consistently favors 8–15 slides
- Ratio: any page size works; square (1:1) and portrait (4:5) occupy the most feed space — export pages at 1080 × 1080 or 1080 × 1350
- File size: keep it under 100 MB, and practically under 10 MB so mobile readers aren't waiting on slide 3
- Text: LinkedIn renders the PDF, so live text stays sharp at any zoom — export text as text, not as flattened screenshots, and your fonts survive because PDF embeds them
Design rules from the format itself
Because slides are PDF pages, the reader's swipe is a page turn, and the design conventions follow:
- Slide 1 is a cover, not content. The feed shows it cropped and small; a big promise in large type outperforms a busy layout.
- One idea per page. Readers swipe in under 2 seconds per slide; a page that needs 10 seconds breaks the rhythm.
- Bottom-right is the "next" zone. A pointer, arrow, or unfinished sentence there pulls the page turn.
- The last page is a CTA. It's the only slide guaranteed a pause.
Where splitting comes in
Real carousel workflows constantly need PDF surgery, and this is exactly what a PDF splitter is for:
- Reuse part of a deck. Your 40-page webinar deck contains a perfect 10-slide story on pages 18–27. Split
18-27into its own file and post it — no design work. - Serialize a long document. A 45-page report becomes a 4-week content calendar: split every 12 pages and you have four carousels queued.
- Trim a designed carousel. Analytics show drop-off at slide 9? Split
1-8, repost the tighter cut, and test. - A/B test covers. Split off
2-12, then merge behind two different cover pages.
Each split preserves the original's fonts, vectors, and links, because pages are copied, not re-rendered.
The image-first alternative
Designing in Figma or Canva as one tall or wide canvas? Slice the export into square tiles with the image splitter, then compile tiles to PDF — though exporting a multi-page PDF directly from the design tool keeps text selectable, which LinkedIn rewards with sharper rendering and accessibility.
Both splitters run entirely in your browser, so client decks and unpublished reports stay on your machine while you cut them down to carousel size. Start with the PDF splitter.