Imagegptgpt-image-2-text-to-image

Gpt Image 2 Text To Image

Generate high-quality images from text prompts using GPT Image 2, supporting up to 20,000 character prompts for detailed and precise image creation.

Open in Image Lab Browse all models~28 credits per run

GPT guideVerified

OpenAI's image lines. GPT Image 1.5 (text-to-image) and GPT Image 2 (text-to-image + image-to-image with up to 16 references) are the flagships. GPT-4o Image is the multimodal-chat variant.

About this variant

Newer OpenAI image model — supports 20k-char prompts.

Strengths

Best instruction-following + scene composition in our image catalog.
GPT Image 2 accepts up to 16 reference images for style transfer / editing.
Reliable for typography (better than Midjourney, behind Ideogram).
Multimodal understanding — interprets reference images at semantic level, not just visual.

Weaknesses

Premium pricing.
Filters reject more prompts than competitors.
Stylization is OK but not its strength.

Best for

Complex scene composition with explicit positioning
Brand-consistent work via reference packs (16 images)
Diagrams, infographics, slides with text

Prompting tips

Be explicit about subjects, positions, what's where.
Long prompts work — GPT 2 supports up to 20,000 characters.
Reference packs: include shots of the SAME subject from multiple angles for best consistency.

Parameters

num_images
int
Number of images to generate
default: 1
range: 1 … 4
output_format
string
Output format for the images
jpegpngwebp
default: png
image_size
string
The size of the generated image. Supports preset names, explicit {width, height}, or 'auto' to let the model pick the best size. Concrete sizes must have both dimensions as multiples of 16, max edge 3840px, aspect ratio <= 3:1, total pixels between 655,360 and 8,294,400.
autosquare_hdlandscape_4_3portrait_4_3landscape_16_9portrait_16_9
default: landscape_4_3
quality
string
Quality for the generated image. Use 'auto' to let the model pick the best quality for the prompt.
autolowmediumhigh
default: high

You'll need

A text prompt

Try now

More from gpt