Available production model on monalisa-1 · OpenAI
Available nowGPT Image 2 AI Image Generator
A production entry with exposed controls for generation, reference-guided editing, aspect ratio and resolution — on monalisa-1, at a fixed 6 credits per run, whatever you select.
Live AI image generator
Create AI images in monalisa-1 studio with the models available today
The monalisa-1 model is visible but disabled. Every request, price and AI image in monalisa-1 studio is tied to the production model you actually select.
Your AI image appears here
Pick an available model, write a concrete prompt, and every result stays tied to the AI image generator model that really produced it.
What is GPT Image 2?
GPT Image 2 is OpenAI's general-purpose image generation and editing model, offered on monalisa-1 through Kie. Separate text-to-image and image-to-image endpoints sit behind one public model page, so you switch workflow without switching product or price.
It is the broadest entry in the monalisa-1 catalog. Prompts run to 5,000 characters, the image-to-image workflow accepts up to 16 reference images, and output covers 1K, 2K and 4K across eight aspect ratios including auto. If you are not sure which model to open first, this is the safe default.
Two scene entries, one fixed price
Text to image and image to image use different upstream entries while sharing one public family, one validated control set and one price: monalisa-1 charges 6 credits for a GPT Image 2 generation. Nothing about that changes when you attach references or move from 1K to 4K.
The interface only exposes options the provider actually accepts, and it checks aspect ratio against resolution before a task is created. Invalid combinations are disabled in the UI instead of failing upstream after your credits are already gone.
GPT Image 2 specification sheet
Every value below comes from the same monalisa-1 server-side catalog that validates and prices your request, so the page cannot show one number while the generator charges another.
- Resolution
- 1K, 2K and 4K, with ratio-dependent limits
- Aspect ratios
- Auto plus 1:1, 4:3, 3:4, 16:9, 9:16, 2:3 and 3:2
- Workflows
- Text to image and image to image
- Credits
- Fixed at 6 monalisa-1 credits per generation
- Prompt length
- Up to 5,000 characters
- Reference images
- Up to 16 images in the image-to-image workflow
- Best for
- Workflows that need generation, reference editing and explicit resolution control
- Known limits
- Auto is 1K-only; 1:1 does not support 4K
How to generate with GPT Image 2 on monalisa-1
Three decisions stand between a prompt and a finished frame, and the monalisa-1 generator above walks them in this order.
- 1
Pick text to image or image to image
Text to image starts from a prompt alone. Image to image unlocks the uploader and takes up to 16 references, which is the mode for edits, variations and product staging.
- 2
Write a specific prompt
You have 5,000 characters, so spend them on subject, framing, lighting and any wording that has to appear in the image. GPT Image 2 is the entry this site points to when a layout has to carry readable words.
- 3
Choose a valid ratio and resolution
Pick the aspect ratio first, then the resolution — auto stays at 1K and 1:1 stops below 4K. Blocked pairs are disabled before a task is created, so you never pay for a request the provider would reject.
GPT Image 2 vs the other models on monalisa-1
Four models are available on monalisa-1 and they differ in price, reference limit and control surface. Here is where GPT Image 2 wins, and where another entry is the better call.
GPT Image 2 vs Nano Banana 2
Nano Banana 2 costs 15 credits against 6 and puts both scenes behind a single model ID with a PNG or JPG switch. GPT Image 2 is cheaper per run, takes more references (16 against 14) and offers the auto aspect ratio, but it enforces stricter ratio-to-resolution rules.
Nano Banana 2GPT Image 2 vs Z-Image
Z-Image is 4 credits, text to image only, capped at a 1,000-character prompt, with no reference uploads and no resolution control. Use it to sketch, then come back to GPT Image 2 when the idea needs editing, 4K or a specific format.
Z-ImageGPT Image 2 vs monalisa-1
monalisa-1 is an art-first model still in development, with every specification listed as TBA — it cannot be selected or charged. GPT Image 2 is what this site recommends in the meantime, and the two target different work: general-purpose production versus artistic expression.
monalisa-1Where GPT Image 2 fits on monalisa-1
- 1
Campaign artwork with words
Posters, ads and covers where the layout has to hold readable copy in a controlled position.
- 2
Reference-guided edits
Variations, reworked scenes and product staging built from up to 16 uploaded references.
- 3
High-resolution concept art
4K exploration across many visual styles, at the same fixed 6 credits as a 1K draft.
GPT Image 2 FAQ
Does GPT Image 2 support reference images?
Yes. Switch to image to image and the uploader accepts up to 16 references before you generate.
Why are some resolution choices disabled?
The provider rejects specific aspect-ratio and resolution pairs — auto only runs at 1K, and 1:1 does not support 4K. The interface disables them before a task or a charge is created.
How are GPT Image 2 credits calculated?
The server resolves the fixed 6-credit price from the same catalog the interface reads. Prices sent by the client are ignored, and the cost does not move with resolution or reference count.
How long can a GPT Image 2 prompt be?
Up to 5,000 characters, the longest allowance in this catalog alongside the Nano Banana entries. Z-Image, by comparison, stops at 1,000.
Which aspect ratios does GPT Image 2 support?
Auto, 1:1, 4:3, 3:4, 16:9, 9:16, 2:3 and 3:2. Auto is the only one locked to 1K output.
Can GPT Image 2 produce 4K images?
Yes — 1K, 2K and 4K are all selectable, except where the chosen ratio restricts it. The credit cost is identical at every resolution.
Is GPT Image 2 free to use here?
No. Each generation costs 6 monalisa-1 credits and needs a signed-in account with a sufficient balance, which is checked before the task is sent upstream.
Should I use GPT Image 2 or Nano Banana Pro?
Start with GPT Image 2 at 6 credits. Run Nano Banana Pro at 45 only when you want a matched comparison on the same prompt and have decided the extra cost is worth testing.
Generate or edit with GPT Image 2 on monalisa-1
Write a prompt, choose a valid ratio and resolution, and generate on monalisa-1 at a server-verified 6 credits.
Create with GPT Image 2