Image AI

GPT Image 2

Generate images with GPT Image 2 using text prompts, input images, quality levels, output formats, moderation controls, and up to 10 images per request.

GPT Image 2

Generate images

Cost: 16 credits

Result

Generated images will appear here.

Examples (copyable)

Use these as starting points. Keep the exact subject terms you care about, then adjust style and constraints.

Text-to-image (product photo)

Studio product photo of a minimalist ceramic mug on a wooden table, soft window light, shallow depth of field, realistic shadows, clean background, natural color.

Image edit (background swap)

Keep the subject exactly the same. Replace the background with a bright modern kitchen. Match lighting direction and color temperature. Preserve texture, edges, and proportions.

Style transfer (still realistic)

Portrait photo, realistic skin texture, subtle film grain, natural lighting, no beauty smoothing, accurate facial proportions, editorial look.

Settings guide (how to choose)

1 image, quick draft
Start with low/medium quality + 1 image. Iterate prompt first, then raise quality only for the final.
Multiple variations (3–10 images)
Keep quality modest. Increase image count to explore ideas, then regenerate the winner at higher quality.
Editing with references
Use 1–3 strong references, and describe what must stay unchanged vs what can change.
Need strict safety
Use the built-in moderation controls; avoid prompts that request copyrighted logos or disallowed content.

Troubleshooting

Text looks wrong or unreadable

Avoid generating critical text inside the image. Generate a clean layout and add text in post, or simplify typography constraints.

Subject identity changes across variations

Repeat identity anchors (hair, outfit, key features) and reduce style adjectives. Use fewer “mood” words.

Output feels over-processed

Add: natural lighting, realistic texture, no beauty smoothing, subtle grain. Avoid “ultra sharp” stacking.

Background swap looks pasted

Add: match lighting direction, match color temperature, consistent shadows, preserve edges.

Too many artifacts

Lower complexity: fewer props, simpler background, shorter prompt. Increase quality only after prompt is stable.

Results drift between batches

Lock key constraints in the first sentence; keep the rest modular. Reuse the same prompt skeleton.

Multi-image output

Generate batches for campaigns, product variants, and creative testing.

Prompt plus image input

Combine written direction with uploaded references for controlled edits.

Quality controls

Choose low, medium, high, or auto based on budget and final-use needs.

FAQ

How many images can GPT Image 2 generate?

This page supports one to ten images per request, with credits calculated from quality and image count.

Can GPT Image 2 edit images?

Yes. Upload input images and describe the edit, style, or new composition you want.

Which quality levels are available?

Low, medium, high, and auto quality are available, with visible credit pricing before generation.

GPT Image 2 Generator | Text to Image and Image Editing | Kling 2.6 Studio