Supports up to 16 reference images and 3000-character prompts for higher-consistency image generation, brand assets, and complex creative workflows
GPT Image 2 is a newly added image generation model on Pearktrue. It supports both text-to-image and image-to-image workflows, accepts up to 16 reference images, and supports four practical aspect ratios: auto, 1:1, 9:16, and 16:9. It is especially useful for users who struggle with character inconsistency, lost product details, mismatched visual styles, and prompts that are too complex for shorter models.
Upload up to 16 reference images for character consistency, product detail retention, multi-element fusion, and style alignment.
Supports both text-to-image and image-to-image for fresh generation, edits, expansions, and style improvements.
Supports auto, 1:1, 9:16, and 16:9 for e-commerce covers, social creatives, vertical posters, and widescreen visuals.
Describe complex creative direction, camera language, visual constraints, and brand requirements in a single prompt.
Text-to-image costs 5 credits and image-to-image costs 8 credits, making planning easier for teams and batch workflows.
Well suited for brand visuals, e-commerce marketing, character design, poster creation, and multi-image fusion tasks.
GPT Image 2 FAQ
Global Reviews
"The multi-reference workflow is the biggest win for me. Character sheets that used to take multiple rounds now land much closer on the first try."
Brand Designer
"It's a huge time-saver for product covers and campaign posters. I can combine product shots, background style, and brand colors much more reliably."
E-commerce Operator
"The 3000-character prompt limit is genuinely useful. We can include campaign goals, composition rules, and brand tone in one generation request."
Content Marketing Manager
"I use it most for character consistency and series artwork. The 16-reference limit cuts down rework dramatically."
Independent Creator
