ImageAd creativesIntermediateGenerates an image

Before-and-after comparison visual

A split-frame comparison for products where a visible difference is the benefit, with guardrails so the 'after' does not overpromise.

For: Performance marketers, Creative strategists, E-commerce marketers · Works with: OpenAI GPT Image, Google Gemini image generation (Nano Banana), FLUX (Black Forest Labs), Any image generator

Your prompt

Create a 1:1 split-screen comparison image divided by a clean vertical line, for an enzyme stain remover for carpets aimed at pet owners. Left half ("before"): a light beige carpet with a visible dark stain, shot in flat, neutral, slightly dull light. Right half ("after"): the same carpet section in the same position, angle and framing, but showing the same carpet with the stain mostly faded, in brighter and cleaner light. Consistency: identical camera angle, crop, background and object placement on both sides so the only visible difference is the result of using the product. Realism: the improvement must be plausible and modest, not exaggerated or impossible. Show a lightened stain, not a perfectly new carpet. Constraints: no text, labels, arrows or logos; no people unless specified: no people.

Customize

The fields start with example values so you can see how the prompt works. Everything stays in your browser.

What is advertised.

Who it is for.

Overall frame.

The thing that changes.

The starting state.

A realistic, modest improvement.

Keep the claim honest.

Whether anyone appears.

Example of what to expect

Illustrative only. It describes the kind of result this prompt aims for; real output varies by tool, model and run.

A square split image: on the left a beige carpet with a dark stain in dull light, on the right the same carpet with a faded stain in brighter light, identical framing on both sides.

Expected format: One split-frame image. Check both halves are the same scene.

How to use it

  1. Make sure the claim matches what your product really does; the image should not promise more.
  2. Regenerate if the two halves differ in angle or objects.
  3. Check the ad platform's rules on before/after imagery for your category.

Limitations

  • Many ad platforms restrict before/after imagery in health, weight-loss and cosmetic categories; check current policy first.
  • AI images are illustrations, not evidence. Do not present them as real customer results.

Platform notes

OpenAI GPT Image

Official docs read · checked 2026-10-11

Aspect ratio is set with the size parameter, not in prose. Use medium or high quality when the image contains text or fine detail.

General notes on OpenAI GPT Image
  • OpenAI's guide covers gpt-image-2 (recommended), gpt-image-1.5, gpt-image-1 and gpt-image-1-mini.
  • Aspect ratio is set through the size parameter, not in prose. Quality is low, medium or high; medium or high is advised for small or dense text.
  • Put literal text in quotation marks, and when editing say what must stay unchanged.

Google Gemini image generation (Nano Banana)

Official docs read · checked 2026-10-11
General notes on Google Gemini image generation (Nano Banana)
  • Google's docs list several Nano Banana models (2.1, 2 Lite, 2, Pro) and recommend 2.1 for new projects as of the verification date.
  • Supported aspect ratios include 1:1, 4:5, 9:16, 16:9 and 21:9; sizes 1K, 2K and 4K. Up to 14 reference images per request.
  • All generated images carry a SynthID watermark.

FLUX (Black Forest Labs)

Official docs read · checked 2026-10-11

FLUX.2 does not use negative prompts, so every constraint here is written as something to do rather than something to avoid.

General notes on FLUX (Black Forest Labs)
  • BFL's official guidance for FLUX.2: no negative prompts (describe what you want), write sentences rather than keyword lists, specify lighting, quote text to render, and pair colour names with hex codes.

Any image generator

Official docs read · checked 2026-10-11
General notes on Any image generator
  • Written as natural-language descriptions that the verified image models accept. Syntax for aspect ratio, quality and negative prompts differs by tool, so set those in your tool's own controls.

Source, license and attribution

Origin
Original by MarketerTools
Publisher
MarketerTools
License
Original work by MarketerTools, free to copy and use. Informed by the linked documentation; no third-party text is reproduced.
Platform assumptions checked
2026-10-11

Documentation and references behind this prompt

These informed the structure and the platform notes. A reference is not a license, and no third-party prompt text is copied here.

Spotted an attribution error, or are you a source owner with a request? Use the chat button at the bottom right and quote prompt ID img-006. We will correct or remove it.

  • ImageIntermediateGenerates an image

    Static ad image: product hero with headline space

    A single-product advertising image with a deliberately empty zone for headline and CTA, so the designer or ad platform can add copy after generation.

    OpenAI GPT Image · Google Gemini image generation (Nano Banana) · FLUX (Black Forest Labs) +2

  • ImageIntermediateWrites a set of prompts

    Creative test set: change one variable at a time

    Builds a fixed base prompt and a controlled list of variants that change exactly one element, so a creative test isolates what actually drove results.

    Any capable chat model (ChatGPT, Claude, Gemini, others)

  • ImageIntermediateGenerates an image

    Meta feed lifestyle ad (4:5)

    A people-in-context lifestyle image for Facebook and Instagram feed, built to read at thumbnail size and to leave room for a short overlay.

    OpenAI GPT Image · Google Gemini image generation (Nano Banana) · FLUX (Black Forest Labs) +1

  • ImageIntermediateGenerates an image

    Vertical Story and Reel ad background (9:16)

    A 9:16 background built around the zones that app interfaces cover, with the subject placed between the top and bottom UI bands.

    OpenAI GPT Image · Google Gemini image generation (Nano Banana) · FLUX (Black Forest Labs) +1