Skip to content

LearnTutorials

Make an infographic with ChatGPT Images 2.6

Turn 3 to 7 facts into one clean, shareable infographic with ChatGPT Images 2.6: set the layout, fix every word and save it in under 20 minutes.

By Recaply EditorsBeginner17 min6 stepsUpdated

A hand lines up three blank paper tiles in a neat column, each with a dot on its left, beside a pencil

What you need

  1. Step 1: Write every word first

    3 min

    The model draws text best when you hand it every word, so settle the words before you open ChatGPT.

    1. Pick one idea for the whole graphic, such as "5 ways to cut meeting time".
    2. Write a headline of 8 words or fewer.
    3. Write 3 to 7 points, each under 12 words, plus a one-line source.
    4. Check every number against its source now. The image will only repeat what you give it.

    You should see a short block of text: a headline, your points and a source line.

    Too many points to fit?

    Keep the 5 strongest, or split them into 2 infographics. Short labels come out sharper than long sentences.

  2. Step 2: Pick the shape and the model

    2 min

    The right shape fits where you'll post it, and thinking plans a busy layout before it draws.

    1. Open a new chat at chatgpt.com and sign in.
    2. Choose the shape: 4:5 for a feed post, 9:16 for a story, 1:1 for most posts, 16:9 for a slide.
    3. On Plus or Pro, pick High in the model picker in the message box. That turns on images with thinking.

    You should see a new chat with your plan's model picked, ready for the prompt.

    No High option in the model picker?

    Images with thinking needs Plus, Pro or Business. On Free or Go, skip it. The prompt in the next step still works.

  3. Step 3: Send the layout prompt

    4 min

    Naming the layout, colors and type up front gets you a usable first draft.

    1. Paste the prompt below and swap in your shape and your text from step 1.
    2. Keep the quote marks around each piece of text, so ChatGPT renders it word for word.
    3. Send it and wait. A dense image can take a minute or more.
    Your prompt
    Create a flat vector infographic, 4:5 portrait.
    
    Layout: [numbered list / side-by-side comparison / timeline], top to bottom.
    Headline at the top, the largest text: "[HEADLINE]"
    Points, each in its own block with a simple icon on the left:
    1. "[POINT 1]"
    2. "[POINT 2]"
    3. "[POINT 3]"
    Footer, small: "Source: [SOURCE]"
    
    Colors: [dark navy] background, white text, [orange] accents. Solid fills, no gradients.
    Type: bold sans-serif headline, clean sans-serif body, high contrast.
    Render the text exactly as written, once each. No other text, no watermark, no logos.

    You should see a first draft with the headline on top and each point in its own block.

    It came out as a photo or a 3D scene?

    Start the prompt with "Flat vector style, minimal infographic, no photos" and send it again.

  4. Step 4: Check every word

    3 min

    A typo in a shared graphic is hard to take back, and image models can still slip a letter.

    1. Zoom in and read each word against your text from step 1.
    2. Check that the headline is the biggest text and each block stands apart.
    3. Fix text first, one message per fix: quote the wrong text and the right text.

    You should see every word on the image matching your text.

    A fixed word breaks again after the next edit?

    Spell the word letter by letter in your request. Add "keep all other text exactly as it is" to every edit after that.

  5. Step 5: Polish the layout

    3 min

    Precise follow-ups beat starting over, since Images 2.6 holds details steady across edits.

    1. Select the image to open the editor, choose Select, and highlight the area to change.
    2. Describe one change, such as "make the headline larger" or "left-align the points".
    3. For whole-image changes, like a new accent color, just ask in the chat.

    You should see your infographic with only the part you asked about changed.

    The edit spilled outside your selection?

    Selections aren't always exact. Choose Undo, select a smaller area, or name the spot in words, such as "the third point only".

  6. Step 6: Save it and keep the prompt

    2 min

    A saved prompt means your next infographic starts from a layout you already trust.

    1. Open the image and select Save to download it.
    2. Need a story version too? Open it in the editor, choose Aspect ratio, and pick a taller shape.
    3. Copy your prompt into a note. Next time, swap in new text and keep the layout and color lines.

    You should see the image on your device, plus a prompt you can reuse.

    Every image you make also stays under Images in the ChatGPT sidebar.

You get

A finished, on-brand infographic ready to post, and a prompt template for the next one

Did this guide work for you?

Sources