AI Image Tools

GPT Image 2 and 2.5

OpenAI's image models, offered here in four tiers: GPT Image 2 at 2K and 4K, and the newer 2.5 at 2K and 4K. They are the models to pick when the picture has to contain words someone can read.

tiers to choose from
4tiers to choose from
credits, by tier
7–20credits, by tier
maximum resolution
4Kmaximum resolution
measured PNG at the 2K tier
2048²measured PNG at the 2K tier
GenerateRestore & EnhanceEditCreatePhoto Effects
0/300
13 credits

Sign in to start — new accounts get 10 free credits, no card.

Example prompts for GPT Image

The same prompts are behind the “Try an example” button in the input box; each click loads the next one.

  • Product shot

    Matte black ceramic mug on a pale linen surface, soft window light from the left, shallow depth of field, editorial product photography

  • Portrait

    Close-up portrait of an older woman laughing, natural window light, film grain, 85mm lens, shallow depth of field

  • Landscape

    Misty pine forest at dawn, low fog between the trunks, cold blue light, wide shot

Made with these models

Images generated on this site with the same models.

  • Isometric illustration of a tiny woodworking workshop, generated with the image models on this site
  • Overhead flat lay of a tidy desk, generated with the image models on this site
  • White sneaker on a studio backdrop, generated with the image models on this site

Why GPT Image on AI Image Tools

  • Text inside the picture

    Signage, packaging, book covers, UI mockups. Put the exact words in quotes and these are the models most likely to render them as letters rather than as shapes that resemble letters.

  • 2.5 follows longer instructions

    The newer generation holds on to more of a complex prompt. If a detailed instruction keeps getting partly ignored on the cheaper tiers, this is the one to try.

  • Real 4K when you need it

    The 4K tiers cost more because they produce more. Our measured output at the 2K tier was already 2048 × 2048 PNG.

  • Reads several reference images

    Up to 16 input images, which makes these tiers usable in Virtual Try-On and any other multi-image workflow.

Every image model in one place

Pick any of these from the Model menu above; the price is shown before you generate. Each name links to what that model is good at. Compare every tier.

  • Z-Image Turbo

    Alibaba · cheapest and fastest, text-to-image only

    1 credits per image

  • Seedream 5 Pro · 1K

    ByteDance · same model as 2K at half the price

    9 credits per image

  • GPT Image 2.5 Flare · 2K

    OpenAI · same price as Sunburst, tuned for speed and volume

    13 credits per image

  • Qwen Image 3 · 2K

    Alibaba · strong at readable text inside the image

    6 credits per image

  • Nano Banana

    Google Gemini 2.5 Flash · fastest, cheapest

    5 credits per image

  • Nano Banana 2.1 · 1K

    Google · newest Nano Banana, sharper text, banner ratios

    10 credits per image

  • Nano Banana 2.1 · 2K

    Google · same model at 2K

    15 credits per image

  • Nano Banana 2.1 · 4K

    Google · for print or heavy cropping

    23 credits per image

  • Ideogram V3 · Turbo

    Ideogram · text inside the image, fastest tier

    5 credits per image

  • Ideogram V3 · Balanced

    Ideogram · text inside the image, posters and logos

    9 credits per image

  • Ideogram V3 · Quality

    Ideogram · text inside the image, highest detail

    13 credits per image

  • FLUX.1 Kontext

    Black Forest Labs · fastest, about 5s per image

    5 credits per image

  • Seedream 5 Lite

    ByteDance · good at Chinese text and posters

    7 credits per image

  • Seedream 5 Pro · 2K

    ByteDance · most photorealistic, highest detail

    18 credits per image

  • GPT Image 2 · 2K

    OpenAI · good with text in images, half the price of 2.5

    7 credits per image

  • GPT Image 2 · 4K

    OpenAI · for print or heavy cropping

    10 credits per image

  • GPT Image 2.5 · 2K

    OpenAI · newest, better instruction following

    13 credits per image

  • GPT Image 2.5 · 4K

    OpenAI · newest, highest resolution

    20 credits per image

GPT Image specifications

Models
GPT Image 2 (2K, 4K) and GPT Image 2.5 (2K, 4K), OpenAI
Modes
Text to Image and Image to Image
Resolution
2K or 4K depending on the tier; measured 2048 × 2048 PNG at 2K
Sizes
1:1, 16:9, 9:16, 4:3, 3:4
Output
PNG, saved to My Creations
Cost
7 credits (2 · 2K), 10 (2 · 4K), 13 (2.5 · 2K), 20 (2.5 · 4K)
Plans
Both 2K tiers on every plan, including free. Both 4K tiers are on paid plans
Time
About a minute on an edit in our runs, the slowest of the tiers we measured
Free plan
10 credits on sign-in; free downloads carry a small watermark

How to use GPT Image

  1. Step 1

    Start on GPT Image 2 · 2K

    Seven credits. Enough to find out whether the model understands the brief.

  2. Step 2

    Move up only for a reason

    2.5 for a long instruction it keeps dropping; 4K for print or heavy cropping. Otherwise the cheaper tier is the same picture.

  3. Step 3

    Check the lettering at full size

    Text is the reason to use these models and also the first thing to go wrong. Zoom in before you download.

How to write a prompt for GPT Image

  • Quote the exact words

    "a sign reading 'OPEN DAILY 7-3'". Unquoted text gets paraphrased or invented.

  • Keep the copy short

    A few words render reliably. A paragraph does not, on any model here.

  • Say where the text sits

    "Centred on the awning", "along the bottom edge". Placement is part of the picture, not an afterthought.

  • Spend the tier on the right thing

    4K buys pixels, not comprehension. If the model is misreading the brief, go to 2.5 rather than to 4K.

Pricing

Every tool draws from one balance.

40% OFF

No hidden fees • Cancel anytime • Unused credits roll over

Basic

16% OFF - Save $18.8

$8.3/month

$9.9/month

$100/year billed yearly

480images / year

$0.21 per image

on Nano Banana

  • 2,400 Credits renew yearly
  • Or 120 images in 4KGPT Image 2.5 · 4K
  • All 26 tools, one balance
  • All 18 models, including the 4K tiers
  • Download without watermark

Pro

20% OFF - Save $70.8
Popular

$24.0/month

$29.9/month

$288/year billed yearly

1,920images / year

$0.15 per image

on Nano Banana

  • 9,600 Credits renew yearly
  • Or 480 images in 4KGPT Image 2.5 · 4K
  • All 26 tools, one balance
  • All 18 models, including the 4K tiers
  • Download without watermark

Max

40% OFF - Save $238.8

$30.0/month

$49.9/month

$360/year billed yearly

4,320images / year

$0.083 per image

on Nano Banana

  • 21,600 Credits renew yearly
  • Or 1,080 images in 4KGPT Image 2.5 · 4K
  • All 26 tools, one balance
  • All 18 models, including the 4K tiers
  • Download without watermark

Swipe to compare plans →

GPT Image: frequently asked questions

What is the difference between GPT Image 2 and 2.5?

2.5 is the newer generation and holds on to more of a long instruction. Both are offered at 2K and 4K. Version 2 at 2K costs 7 credits; 2.5 at 2K costs 13.

Which tier should I start on?

GPT Image 2 · 2K at 7 credits. Move to 2.5 when the model keeps dropping part of your instruction, and to 4K only when you need the pixels for print or cropping.

Are these the best models here for text in an image?

Yes, alongside Seedream for poster layouts. Put the exact words in quotes and keep them short.

Is 4K worth double the credits?

The 4K tiers are on paid plans. Even then, only if you need the resolution. The picture is the same picture; you are paying for pixels. For a larger file from a cheaper tier, run the 2K output through the Image Upscaler for 2 credits.

What happens if a run fails?

You are only charged when an image comes back. A failed or timed-out run costs nothing.

Put readable words in the picture

GPT Image 2.5 · 2K is preselected above. Switch tiers in the model menu.

Try GPT Image nowSee pricing

The other models

All of them are in the same dropdown, and one credit balance covers every one.

Compare every tier side by side