How many panels can it handle?

Four comfortably, five at a push. Beyond that the faces get too small to carry expression, and expression is what makes a wordless comic readable.

Should I put the dialogue in the prompt?

No. Generate clean panels and add lettering in an editor. Four small text areas in one image is the least reliable thing you can ask for.

Will the character stay the same across all four panels?

Usually close, not guaranteed. Within one generation consistency is decent. If it has to be exact, establish the character in a separate image first and pass it as a reference.

Can I do a horizontal strip instead?

Yes — change the ratio and say "horizontal strip" instead of "vertical reel". The panel-per-line structure is unaffected.

GPT Image 2.5 prompt — a four-panel comic, one beat per panel

Write the panels, not the story. One visible action each, numbered.

A four-panel vertical comic of a pet left home alone, from the owner leaving to the pet posing innocently at the door

Settings the Turn a story into a four-panel comic prompt ran at
Mode	Model	Quality	Size	Ratio	Credits
Text to Image	Think	Standard	1K	2:3	6 cr

Published by OpenAI as an example for this model — not generated on this site. Model: gpt-image-2.5-sunburst. Source: OpenAI The published example ran at 1024x1536, a legal size we do not currently offer; the frame here is the same shape at our nearest size.

A pet left alone in the house, in four beats: the goodbye, the silence, the takeover, the innocent pose at the door. Multi-panel output works when the prompt is already a shot list — each panel gets its own line, and each line describes one thing a reader could see happen. Narrate instead and the model has to do the editing, which is where panel comics usually fall apart.

In full

The prompt, nothing cut

Create a short vertical comic-style reel with 4 panels.
Panel 1: The owner leaves through the front door. The pet is framed in the window behind them, small against the glass, eyes wide, paws pressed high, the house suddenly quiet.
Panel 2: The door clicks shut. Silence breaks. The pet slowly turns toward the empty house, posture shifting, eyes sharp with possibility.
Panel 3: The house transformed. The pet sprawls across the couch like it owns the place, crumbs nearby, sunlight cutting across the room like a spotlight.
Panel 4: The door opens. The pet is seated perfectly by the entrance, alert and composed, as if nothing happened.

639 characters

Run it here

This exact prompt, at these exact settings

The mode and all five settings are locked to this template, so there is nothing to line up by hand. Press Run this prompt above and the panel opens filled in — prompt, mode, quality, size and frame — and your first image needs no account. If your plan sits below what this ran at, the line under the panel says so.

This example was published at Think · Standard · 1K. A free run is Fast · Standard · 1K, so expect less fine detail and softer small text. The composition is the same; the finish is not.

3 levers

What is safe to change

Most libraries publish this list and stop, which is why so many copied prompts come back worse than the original: the swappable nouns are the safe half. The three sections under it are the other half — the clauses that are doing the work, the ones that break it if you touch them, and why it is written in this order.

The panel numbering
"Panel 1:" … "Panel 4:" is the layout instruction and the sequencing instruction in one. Change the count and the grid follows. Four is a good default because it fits a vertical frame at a readable size.
One beat per line
Each panel line contains exactly one change of state. That is the discipline the whole template rests on — a panel describing two things happening produces a panel where neither reads.
The physical tells
"Paws pressed high", "posture shifting", "seated perfectly". Comics carry emotion through body language because there is no soundtrack, so describe the body, not the feeling. "The pet is sad" gives you much less than "paws pressed high".

Three ways to break it

Each one is a real failure with a reason attached. A rule without a reason is not usable.

01Adding dialogue or captions
Four panels means four separate small text areas, which is the hardest text-rendering job there is. Generate the artwork with room for lettering and add the words in an editor — this is the standard workflow for a reason.
02Going past four or five panels
Panel count divides a fixed canvas. At six the faces are too small to carry expression, which is the only thing making the story legible, and at eight you have a grid of thumbnails.
03Describing the character once and assuming it holds
Consistency across panels within a single generation is decent but not guaranteed. If the character must be recognisable, establish it in its own image first and use that as a reference — that is what the children's-book template is for.

Why it is written this way

Inside the Turn a story into a four-panel comic prompt

Multi-panel generation is one of those jobs that looks like it should be hard and is mostly a writing problem. The model can draw four related pictures in a grid. What it cannot do is decide where the cuts go, and that decision is the entire craft of a comic.

So this prompt does not tell a story. It hands over a shot list. Every panel line is a single frozen instant with one visible change in it: a person exits, a door shuts, an animal is on the furniture, an animal is by the door. Nothing in any line requires knowing what happened in a different line to be drawable. That property is what makes the grid come out coherent.

The joke — and it is a well-constructed joke — lives in the gaps between panels rather than in any panel. Nothing in panel 3 explains the crumbs, and nothing in panel 4 acknowledges panel 3. Comics work this way; readers do the connecting. What that means for prompting is that you should not try to draw the connection. A line like "the pet realises the owner is coming back and rushes to the door" is asking for two moments in one frame, and the model will either pick one or blend them into something confused.

The second craft note is that all the emotion is physical. "Eyes wide, paws pressed high" in panel 1; "posture shifting, eyes sharp with possibility" in panel 2; "seated perfectly, alert and composed" in panel 4. No feeling word appears anywhere. This is not stylistic preference — a model rendering "the pet feels guilty" has to invent a physical expression of guilt, and the one it invents will be generic. Describing the pose directly gets you the specific version you had in mind.

On text: leave it out. Four panels means four small lettering areas, which stacks the hardest text job onto the hardest layout job. Ask for clean panels and letter them afterwards.

4 questions

Turn a story into a four-panel comic — common questions

How many panels can it handle?: Four comfortably, five at a push. Beyond that the faces get too small to carry expression, and expression is what makes a wordless comic readable.
Should I put the dialogue in the prompt?: No. Generate clean panels and add lettering in an editor. Four small text areas in one image is the least reliable thing you can ask for.
Will the character stay the same across all four panels?: Usually close, not guaranteed. Within one generation consistency is decent. If it has to be exact, establish the character in a separate image first and pass it as a reference.
Can I do a horizontal strip instead?: Yes — change the ratio and say "horizontal strip" instead of "vertical reel". The panel-per-line structure is unaffected.