Find the right AI
English

Which AI should I use to create images?

Most people don't need a separate image app: ChatGPT, Gemini and Meta AI all make and edit images right in the chat. Here's which one to pick for text in images, edits of your own photos, or free everyday use.

Which AI does this best

  1. ChatGPT
    AI assistantFree plan
    #1

    Best at following detailed instructions and getting text in the image right, so it's great for posters, thumbnails and product shots. You edit by just asking. The free plan gives you only a few images a day.

    ✦ Try this prompt
    Create a [square / 4:5 / 16:9] image for [use: Instagram post / YouTube thumbnail / poster]: [describe the scene, subject, setting]. Style: [photo / flat illustration / 3D]. Colours: [colours]. Put the text "[exact text]" in [position], large and easy to read. Leave space at the [top] for a logo. No other text.
    Open ChatGPT ↗
  2. Gemini
    AI assistantFree plan
    #2

    Its Nano Banana model is excellent at editing photos you upload: change the background, outfit or light while keeping the face the same. It's free for everyday use. Text inside images is less reliable than ChatGPT's.

    ✦ Try this prompt
    Here's my photo [attach]. Keep my face, hair and pose exactly the same. Change the background to [new setting], the light to [soft daylight / golden hour], and my outfit to [outfit]. Make it look like a real photo taken on a phone, not an illustration. Then give me a second version in [4:5] for Instagram.
    Open Gemini ↗
  3. Meta AI
    AI assistantFree
    #3

    Free, and Meta's new image model can generate from text or from reference photos and do quick edits. Handy if you're already in Instagram or WhatsApp. Less control over details than ChatGPT or Gemini.

    ✦ Try this prompt
    Create an image of [subject] in [setting], [style: realistic photo / watercolour / cartoon], [mood] mood, [colours]. Use the photo I've attached as a reference for [the product / the person / the style]. Make 4 versions with different angles.
    Open Meta AI ↗
  4. Grok
    AI assistantFree plan
    #4

    Grok Imagine makes bold, stylised images and can turn them into short clips with sound. Image generation is paid-only now, and it's less precise with text and small details.

    ✦ Try this prompt
    Generate an image: [subject], [action], [setting], [style: cinematic photo / anime / poster art], [lighting], [aspect ratio]. Then give me 3 variations with a different colour palette each.
    Open Grok ↗

Specialised tools

  1. Midjourney
    Paid

    Still the go-to for beautiful, artistic images with a strong style. There's no free plan.

    Open Midjourney ↗
  2. Ideogram
    Free plan

    Pick this when your image needs words in it, like a poster, a quote card or a logo idea.

    Open Ideogram ↗

Last updated:

Popular tasks