Models › Image models

GPT Image 2.6

OpenAI’s newest image model, the one in ChatGPT Images 2.6: sharpest text and detail

OpenAI · image model

Default pickFeatured

Try GPT Image 2.6 →

10aspect ratios

up to 16reference images

Standardoutput resolution

Prompt intext to image

GPT Image 2.6 is one of 35 image models Hermoso can render a still ad, product shot or thumbnail with. Name it, or leave the model out and the router picks a sound default. The same catalog is available from the app, the API, the MCP server and the CLI, on one account and one credit balance; the exact price of every render is shown in the app before it runs.

What people make with it

The linked formats are creative recipes Hermoso plans and renders for your brand.

Use cases

GIF and SVG

AI GIF generator: GPT Image 2.6 draws every frame of a GIF that loops, from one sentence
AI SVG generator: logos, icons and illustrations as editable SVG files

About OpenAI

OpenAI makes the GPT Image line, the strongest models at rendering legible text inside an image, and the GPT text models.

Specs

Aspect ratios	1:19:1616:93:44:32:33:24:55:421:9
Reference images	up to 16 (edit)
High resolution	No
Model id	`gpt-image-2.6`

Where it fits

Each aspect ratio GPT Image 2.6 renders maps to a placement:

1:1 for Feed posts and carousels
9:16 for Reels, TikTok, Shorts and Stories
16:9 for YouTube, in-stream and display
3:4 for Pinterest and tablet feed
4:3 for Presentations and display
2:3 for Pinterest pins
3:2 for Photography and web hero
4:5 for Instagram and Facebook feed
5:4 for Print and gallery crops
21:9 for Cinematic banners and hero films

GPT Image 2.6 compared with other image models

Model	Ratios	Refs	Hi-res
GPT Image 2.6	10	up to 16	No
GPT Image 2.6 Flare	10	up to 14	No
GPT Image Mini	3	up to 14	No
FLUX1.1 Ultra	9	up to 14	No
Nano Banana 2.1	14	up to 14	Yes
Grok Imagine 2.0	15	up to 14	Yes
Ideogram v4 Fast	10	up to 14	No

See all 35 image models side by side.

Just ask for it

In Claude, ChatGPT, Cursor or any assistant connected to Hermoso, name the model in plain English. Or pick it from the model menu in the app.

Make a 4:5 product shot of our olive oil on marble with GPT Image 2.6.

Use GPT Image 2.6 for three static ads with the headline "Pressed this week" and schedule them.

Put our bottle on a beach with GPT Image 2.6, keep the label exact.

Getting the best out of GPT Image 2.6

Put any on-image text in quotes and say where it sits. Keep it under six words.
Attach the packshot and ask for it unchanged; describe the scene around it, not the product.
State the placement's ratio (4:5 for feed, 9:16 for Stories) so nothing gets cropped later.
Name the light and the surface (morning light, marble) instead of adjectives like premium.

Make an image ad with GPT Image 2.6. Start free with the full studio and up to 10 posts a month.

Try GPT Image 2.6 → See pricing