2026-07-25

How to Make a Meme: Text, Fonts, and Export Settings

Make a meme that stays readable everywhere. Why Impact works, when captions belong outside the image, and which export settings survive social platforms.

How to Make a Meme: Text, Fonts, and Export Settings

Last updated: July 25, 2026

A meme is a picture plus a caption, which sounds like it should take five seconds. Most of them do. The ones that come out badly usually fail for the same handful of reasons: the caption is unreadable against the image behind it, it runs off the edge on a phone, or the file gets mangled by whatever it was posted to.

This guide covers the small number of decisions that actually determine whether a meme reads well, and the export settings that keep it looking the way you made it.

Quick answer: what makes a meme readable?

White text in a heavy condensed font with a thick black outline, uppercase, placed at the top and bottom of the image. That combination has been the default for fifteen years because it is genuinely the most robust: it stays legible over light backgrounds, dark backgrounds, and busy ones, at any size, without you having to think about the image underneath.

If the caption is long, move it out of the picture and onto a solid bar above or below instead of shrinking the text. The Imagic AI Meme Generator does both layouts, and runs in the browser with no signup and no watermark.

Why Impact, specifically

Impact is not a good font. It is an effective one, for reasons that have nothing to do with taste:

  • It is very heavy. Thick strokes survive an outline being drawn around them. A light font with a 4-pixel black stroke turns into a black smudge.
  • It is condensed. More characters fit per line, so a caption is less likely to wrap awkwardly or need shrinking.
  • It has tight spacing and simple shapes. It stays readable when scaled down to a thumbnail in a feed.
  • It ships everywhere. It has been bundled with Windows and macOS for decades, so a meme made with it looks the same on the machine that renders it.

The visual convention that grew around it — white fill, black outline, all caps — is doing the real work.

That outline is what makes the text readable regardless of what is behind it. Without it, white text vanishes over a bright sky and black text vanishes over a shadow. It is the same principle behind accessible contrast guidance: the WCAG contrast requirements exist because text over an uncontrolled background is unreliable, and an outline is how you control it when you cannot control the background.

The stroke is drawn with rounded joins rather than sharp ones. On a heavy font at large sizes, mitred joins produce visible spikes where strokes meet — the Canvas lineJoin documentation shows the difference.

Other fonts are fine when you want a different tone. Just keep the outline.

Font Reads as Good for
Impact Classic meme, loud Reaction images, the default choice
Arial / Verdana Neutral, informational Explainer graphics, screenshots with captions
Comic Sans MS Deliberately unserious Ironic or self-aware jokes
Times New Roman Formal, mock-official Fake announcements, "serious" parody
Courier New Technical, typewritten Developer humour, terminal jokes

Different typeface styles printed side by side, showing how weight and width change caption readability

Inside the image or outside it?

This is the decision that most affects readability, and it comes down to caption length.

Situation Layout Why
Short caption, two to six words On the image Classic look, nothing is covered
Long caption, a full sentence or more On bars above/below Shrinking text to fit is worse than moving it
The subject fills the frame On bars Overlaying would cover the thing being pointed at
Busy or high-contrast background On bars, or a very heavy outline Outlines struggle against extreme detail
The image is the joke, the text is context On bars Keeps the picture intact

Text on bars is sometimes treated as the lesser option. It is not — it is the correct choice whenever the caption is long enough that overlaying it would either cover the subject or force the type down to a size nobody reads on a phone.

Step by step

  1. Pick the image at the largest size you have. You can always scale down. Starting from a small, already-compressed copy means the caption sits on top of visible JPEG artefacts.

  2. Write the caption before you start styling. The length determines the layout, and the layout determines everything else.

  3. Choose the layout using the table above.

  4. Set the text size. Around 10% of the image height is the classic proportion for a caption on the image. Larger reads as shouting, which is sometimes the point.

  5. Check the outline against the lightest and darkest parts of the image. This is the step people skip. A white caption with a black outline over a white wall depends entirely on that outline being thick enough.

  6. Export and look at it small. Most people will see this at thumbnail size in a feed, not full screen.

The Meme Generator wraps long captions and reduces the size automatically so the text always fits the frame, rather than letting it run off the edge.

Close-up of a vintage typewriter sheet, illustrating how heavy letterforms hold up at small sizes

Export settings that survive social platforms

Every platform re-encodes what you upload. You cannot prevent that, but you can avoid handing it something that compresses badly.

Export PNG from the editor. Text with a hard outline is exactly the kind of high-contrast edge that JPEG handles worst — it produces faint ringing artefacts around every letter. Starting from a lossless PNG means the platform's own compression is the only generation of loss.

Do not upscale. If the source image is 600 pixels wide, exporting at 2000 does not add detail; it just makes a larger file that gets compressed harder.

Match the aspect ratio to where it is going rather than letting the platform crop it. A caption that sits near the bottom edge is the first casualty of an automatic square crop. The social media image sizes guide has the current dimensions for each platform.

Compress afterwards if the file is large. A meme with bars is mostly flat colour and compresses very well. Run the PNG through the Image Compressor if you need to get under an upload limit — but compress once, at the end, not repeatedly.

If you are posting to Instagram specifically, check the result against the current Instagram post dimensions before uploading, since a crop applied after the fact will cut into the caption.

Code on a monitor showing high-contrast text edges, the kind JPEG compression handles worst

Common mistakes

Shrinking the text to fit. If the caption does not fit at a readable size, the layout is wrong, not the size. Move it to bars.

No outline. The single most common cause of an unreadable meme. Even a dark caption over a light image needs one, because "light" is rarely uniform.

Re-saving a meme that was already saved as JPEG. Each round of JPEG compression degrades the text edges further. Work from the original image, not from a copy that has already been through a platform.

Very long lines. Even when they fit, a caption running the full width of a wide image is hard to read. Two shorter lines beat one long one.

Assuming the font renders the same everywhere. If you make the image and export it as a picture, you are safe — the type is baked into the pixels. This only bites if you are sharing an editable file.

Frequently asked questions

What size should a meme be? Match the platform, not a universal number. A square around 1080×1080 works almost everywhere and survives most automatic crops. Wider images get cropped or letterboxed depending on where they land.

Is Impact free to use? Impact ships with Windows and macOS and is available to use in documents and images you create on those systems. Embedding the font file itself in software or redistributing it is a different question governed by its licence — but typing a caption with it and exporting a picture is ordinary use.

Can I use any image I find? Making a meme does not automatically make the underlying image yours to publish. Photographs are copyrighted by whoever took them. In practice a great deal of meme culture operates on widely-circulated stills, and enforcement is rare — but "rare" is not "never", and it matters much more if the meme is being used commercially or for a brand. Using your own photo, or one with a permissive licence, removes the question entirely.

Does the generator add a watermark? No. The Meme Generator exports exactly what you see, with no branding added, and does not require an account.

Does my image get uploaded? No. The captions are drawn onto the image by your browser on your own device, and the file is written back out locally. Nothing is sent to a server.

Why is my text getting smaller than I set it? Because it would not otherwise fit. Long captions are wrapped onto multiple lines and reduced until they fit the frame — the alternative is text running off the edge of the image. If you want it larger, shorten the caption or switch to the bars layout.

Summary

The classic meme look — heavy condensed font, white fill, thick black outline, uppercase — is a durable solution to a real problem: keeping text readable over an image you did not design around. Keep the outline, move long captions onto bars instead of shrinking them, export PNG, and check the result at thumbnail size before posting.

Start with the Meme Generator. If you need the caption somewhere other than the top or bottom, Add Text to Photo gives you free placement, and the Image Compressor will get the file under an upload limit if you need it to.

Image credits

Use the free tools while you follow the guide.