Lettering in many languages
Korean, Japanese and English on the same poster or menu, spelled the way you wrote it.
Lettering in many languages, at half the price.
20% off with GPTIMAGE20 GPTIMAGE20 applied
GPT Image 2 is the OpenAI image model behind ChatGPT Images 2.0, released on April 21, 2026. It brought readable text in many languages, flexible layouts and detailed edits, and it is still one of the strongest image models available.
Flare and Sunburst now rank above it and run faster, but it still placed third on LMArena in late September 2026, ahead of every model outside OpenAI. On Fiddl it costs half as much per image, a smart pick for posters, diagrams, comics and big batches.
Korean, Japanese and English on the same poster or menu, spelled the way you wrote it.
Recipe cards, exploded views and labeled diagrams with clean structure. These two were made by Fiddl creators.
Panels with speech bubbles, consistent characters and book covers with real titles.
About 250 images with the Studio pack instead of 125, at the same high quality setting. Slower, but great for big batches.
One try per model with Fiddl's settings, nothing retouched. Times are from our run on October 3, 2026.
All three spelled every line correctly.
Japanese and prices correct on all three.
Very close in style; each kept the fox, the owl and the lantern.
Public images Fiddl creators made with GPT Image 2.
OpenAI's own advice: if GPT Image 2 already does the job, try Flare for speed; if GPT Image 2 falls short on a hard case, try Sunburst. Your Fiddl balance works with all three, so you can compare on your own prompts.
| Model | Flare | Sunburst | GPT Image 2 |
|---|---|---|---|
| Best for | Fast everyday images and edits | Precise edits and polished campaign work | Lettering and layouts on a budget |
| Speed in our test | 27 s | 40 s | 99 s |
| LMArena text-to-image September 24, 2026 | #2 | #1 | #3 |
| LMArena image editing September 29, 2026 | #2 | #1 | #3 |
| Images with the Studio pack | about 125 | about 125 | about 250 |
| Reference images | up to 16 | up to 16 | up to 10 |
| Released |
Condensed from OpenAI's image prompting guide and our own tests.
Launch offer 20% off any pack with code GPTIMAGE20
Code GPTIMAGE20 applied: 20% off every pack
Already on Fiddl? Open GPT Image 2 in the app Use code GPTIMAGE20 on your next pack
OpenAI's image model released on April 21, 2026 with ChatGPT Images 2.0, known for readable multilingual text, flexible layouts and precise edits. On Fiddl it takes text and up to 10 reference images and makes images up to 4K.
Yes, for the right jobs. The 2.5 models rank higher and are two to four times faster, but GPT Image 2 still produced every word correctly in our tests and costs half as much per image on Fiddl.
On LMArena's text-to-image leaderboard of September 24, 2026, GPT Image 2 ranked third and Nano Banana Pro fourteenth; on the image-edit board GPT Image 2 was third and Nano Banana Pro tenth. For a specific look it is still worth trying both; Fiddl also has Nano Banana 2 and Seedream 5 Pro.
About 50 images with the $10 Starter pack and about 250 with the $42.50 Studio pack, from $0.14 per image on the largest pack. 4K images cost a little more. During the launch, code GPTIMAGE20 takes 20% off any pack until November 2, 2026.
OpenAI lists transparent output for GPT Image 2 as a preview, and Fiddl currently returns images with a normal background. Ask for a plain white background, then remove it with Fiddl's background remover.
Yes. Our Seoul poster and Tokyo ramen menu were made on the first try with the Korean and Japanese text exactly as written in the prompt.
There is no free tier for GPT Image 2 on Fiddl. Packs are one-time payments from $10 with no subscription.