Loading GPT Image 2...

Built for serious creators

Three things that make GPT Image 2 different

Razor-sharp output up to 4K

Generate at 1K, 2K, or 4K. Landscape, portrait, or square — every aspect ratio supported.

Text-to-image and image editing in one canvas

Drop a reference image, then describe the change in plain language. No layers, no masks, no fuss.

One-click hand-off to Seedance video

Turn any generated image into a video without re-uploading. The frame flows straight to the Seedance workspace.

Three steps to a finished image

No prompt syntax to memorize. No settings to tune.

1

Describe what you want

Use natural language. Reference an image if you want to edit instead of starting from scratch.

2

Pick resolution and aspect ratio

1K for quick drafts, 4K for hero shots. Choose any ratio.

3

Generate, then animate

Watch the result land in seconds. Download it, or send it straight to Seedance to turn it into a video.

Frequently asked questions

What is GPT Image 2?

A state-of-the-art image model that produces clean, photorealistic results from natural-language prompts. It supports both text-to-image and image editing in one workflow.

Which resolutions and aspect ratios are supported?

Outputs are available at 1K, 2K, and 4K, in landscape, portrait, and square aspect ratios.

Can I edit existing photos?

Yes. Upload a reference image and describe the change in plain language — no masking or layer work required.

How much does each generation cost?

Each generation costs 5 to 24 credits depending on the resolution you pick. Active subscribers automatically pay the lower per-image rate.

Can I turn the generated image into a video?

Yes. Press "Image to Video" on any result to hand the frame off to Seedance and start animating immediately.

Start creating

Sign in to claim your trial credits, or jump straight to the pricing page.