Comic Panel Layout: How Many Panels Per Page, and How to Pace Them

You have the script. You know who is in the scene and what they say. The question that stops most people next is dumber and harder than it sounds: how many panels go on this page, and how big should each one be?

There is no single right number, but there is a working range. This guide covers the practical craft of panel layout: counts that actually work, when to spend a full-width panel, how an establishing shot earns its space, and how the same scene gets carved up differently for a printed page than for a vertical scroll.

What a panel layout actually decides

A panel is a unit of time. The gutter between two panels is the gap the reader fills in, and the size of a panel tells the reader how long to sit in it. That is the whole grammar. Everything else is application.

So a layout is not decoration. It is a set of timing instructions. Six equal panels in a grid read as steady, a little clinical: good for procedure or mounting routine. One panel taking half the page reads as a held breath. Three narrow slivers in a row read as fast cuts. When a page feels wrong and you cannot say why, the usual cause is that the panel sizes are telling the reader a different rhythm than the writing is.

How many panels on a print page: four to six is the working answer

For a paginated comic book, four to six panels per page is the range most pages live in, and it is a good default to build from.

Under four, the page spends its space on very few moments. That is a deliberate choice for a reveal, but a book made of three-panel pages burns through story fast and reads thin. Over six, panels get small, dialogue fights the art for room, and lettering has nowhere to go without covering a face. Seven or eight can work for a montage where each panel carries one short line, but it is a special case, not a default.

A few patterns worth keeping:

• Two rows of three, or three rows of two: the plain workhorse. Steady pace, easy lettering, nothing calls attention to itself.
• A wide panel on top, smaller ones below: establish, then break the moment down. This is the most reusable page in comics.
• Five panels with one oversized: four beats at normal speed plus one that lands. Put the oversized panel where the emphasis actually is, not automatically at the bottom.

The last panel on a right-hand page is doing extra work whether you planned it or not, because the reader has to physically turn to get the answer. Put your question there.

The webtoon page is a scroll length, not a count

In a vertical-scroll strip there is no page, so panel count per page is the wrong unit entirely. The unit is distance. A reader scrolls at their own speed, and the empty vertical space between panels is the pause.

That changes the arithmetic. A tall panel, a long gap, then a small panel is a beat, a silence and a punchline, and it can occupy more screen than six panels of conversation. For an episode of ordinary length you are looking at roughly forty to eighty panels down a strip the reader takes a couple of minutes to travel, with the gaps doing the work page breaks do in print.

If you are choosing between the two formats rather than laying one out, the trade-offs are covered in print comic book vs. webtoon. This guide assumes you have already picked one.

When to spend a full-width panel

A full-width panel is the most expensive thing on your page. It uses the space of three or four normal panels, so it has to buy something those three or four could not.

Good reasons to spend it: a location the reader has not seen and will spend the next ten pages in; a physical scale that only works at size, like the ship or the crowd or the drop; a single emotional beat you want the reader to sit inside rather than pass through; the end of a chapter.

Bad reason, and the most common one: the art came out well. A striking panel is not automatically a big panel. If the moment does not need the time, giving it the time slows the story down for the price of a nice picture.

One per page at most. Two big panels on one page is usually one of them doing nothing.

The establishing shot buys the reader their bearings

Every time the scene changes location, the reader has to rebuild the room in their head. An establishing shot, a wider panel showing where we are before we get close, does that work once so the following five panels do not have to.

Skip it and you get a recognizable failure: the reader follows the dialogue fine but has no idea where anyone is standing, and two pages later a character walks through a door nobody knew was there. The fix is cheap. One wide panel at the top of the scene, and the rest can be close-ups.

It does double duty as a transition. Cutting from a close-up of one face straight into a close-up of another in a different place is disorienting; a wide shot between them tells the reader plainly that we have moved.

One beat per panel, and the thing that breaks when a beat has no place in it

The rule underneath all of this is one beat per panel. If a panel holds two actions, one of them is invisible. Split it.

Here is the part that is specific to generating panels rather than drawing them, and it is worth saying plainly because it costs people real time. AudioProducer generates one panel per story beat, working from the beat text itself. A beat that names a place produces a panel set in that place. A beat with no location noun in it has nothing to anchor to, and the setting drifts. The panel that renders a line of pure dialogue, two characters talking with nothing in the text about where they are, is the panel that comes back with a different room than the one before it.

This is not a flaw you fix panel by panel after the fact. It is an input problem with an input fix: anchor every beat to a place, including the dialogue beats. "She tells him no" becomes "She tells him no, still standing in the kitchen doorway." That one clause is the difference between a scene that holds together and a scene you re-generate twice.

The same discipline applies to the cast, for the same reason and with a different mechanism. Characters are held by reference images rather than by the beat text, which is covered in keeping your characters consistent in an AI comic.

Thumbnail the page before you generate it

A thumbnail is a rough sketch of the page as boxes: no art, just sizes and positions. It takes a minute per page and catches the problems that are expensive later.

Three things it catches that a script does not. Whether the dialogue fits: two characters trading four lines each will not fit in a panel the size you drew. Whether the reading order is unambiguous, which is where an oddly shaped panel gets caught. And whether the page has any shape at all, or is six identical boxes by default.

In practice this means deciding the layout deliberately rather than accepting the first split. The pages and panels are generated with varied layouts, and a page can be re-laid out, merged, reordered, or split after the fact, so a thumbnail is a plan you can actually execute rather than a wish. Where the bubbles land is its own craft, covered in adding speech bubbles and lettering to a comic.

From panels to a finished issue

Layout is the middle of the pipeline, not the end. The chapter becomes pages and panels, the panels get art conditioned on your style and your cast, the bubbles get placed, and the issue exports as a print-ready PDF or a vertical strip. If you are starting further back than layout, making an AI comic from your story covers the whole path, and scripting a webtoon covers the beat and dialogue writing that feeds into these pages.

And because the same source text also carries the performance, the cast you just laid out across your panels can be voiced, with each character on their own voice.

To see your own chapter split into pages and panels, start a project and bring the first chapter. Layout is the part you argue with, and it is easier to argue with a page that already exists. A full index of our comic, webtoon and audio guides lives at all articles.