Gadgetify shares a GPT Image 2 prompt for dual-read hybrid images

On June 19, Gadgetify (@Gdgtify) cited in This one has a lot of potential»—. It's not figure-ground like the Rubin/FedEx illusion prompt: here two subjects coexist in the same image depending on the reading distance, with layers of fine detail and soft tones.

AI Prompt · How to do it

Example posted by Gadgetify on June 19, 2026. @Gdgtify

Steps

1. Open chatgpt.com with GPT Image 2 active.
2. Replace [Subject A] (near reading) and [Subject B] (far reading) with two matching shapes—e.g. «butterfly» + «woman's face»—.
3. Copy the complete prompt below with the weights ::.
4. For a 2×2 grid like the one in the example, add at the beginning: “2x2 grid, four hybrid image pairs.”
5. Try squinting or moving away from the screen to check the second reading.

Prompt to copy

A hybrid image that reads as [Subject A] up close and as [Subject B] from a distance or when squinting ::5 Frequency Logic: [Subject A] lives in the fine sharp detail; [Subject B] lives in the soft blurry low-frequency tones — infer two subjects whose shapes roughly align so both layers register ::5 Composition: Both subjects centered and same scale, edges aligned, neutral background ::3 Material Physics: Crisp high-detail layer overlaid on a soft tonal layer, balanced contrast ::2 Illumination: Flat even light so neither layer dominates ::1 Render Stack: Optical-illusion artwork, MIT hybrid-image study, fine-art photo composite ::2 Negative: [only one subject readable, misaligned subjects, too much detail drowning the far reading, harsh outlines, color clash, text] :: -1

Text in English according to the original publication. The ::n weights prioritize prompt blocks. Results vary by ChatGPT plan and region.

What makes this prompt special

MIT hybrid images superimpose a high-frequency layer (fine detail, visible up close) and a low-frequency layer (blurred shapes, visible when zoomed out). The Frequency Logic block (::5) translates it into AI language: Subject A in sharp strokes, Subject B in soft tones, with shapes that "fit" so that both readings work.

Unlike the binary figure-ground prompt, here there can be color, texture and depth — like a butterfly on a face, an owl made of trees or a skull made of a skyline. The negative prompt penalizes misaligned readings, too much detail that drowns out the distant view, or color clashes.

Relationship with other Gadgetify prompts

Gadgetify has been testing templates with :: weights in GPT Image 2 for days: knolling 2×2, illusions figure-background and World Cup dioramas. This adds MIT hybrid-image logic to the repertoire—ideal for posters, product design, or visual experiments with double reading.

A/B Pair Ideas

• Butterfly ↔ woman's face (as in the example above left)
• Owl ↔ tree forest
• Skull ↔ urban skyline
• Cat ↔ underwater coral reef

Before publishing

Hybrid images may include accidentally recognizable faces or brands. Review the result before using it commercially and respect OpenAI's policies for image generation.