We’re sharing GPT Image 1.5, an image generation and editing model focused on high instruction fidelity, realistic detail, and reliable multi-step image edits.
What GPT Image 1.5 does well:
Strong Prompt Understanding
Handles complex prompts involving layout, style, and object relationships with higher consistency than earlier image models.
Precision Image Editing
Modify existing images—add, remove, or adjust elements—while preserving lighting, composition, and identity across iterations.
Improved Text & Layout Rendering
More reliable generation of readable text, UI-like layouts, posters, and structured visual designs.
Faster Iteration
Optimized generation speed enables rapid experimentation and creative iteration.
Why we built this:
Many image models excel at one-off generations but struggle with control, consistency, and iterative edits. GPT Image 1.5 is designed for creators and builders who need images that stay coherent as ideas evolve.
We’d love feedback on:
How you’d use GPT Image 1.5 in real creative or product workflows
Where image editing still breaks down
What controls or API features would matter most to you
Try it here: https://www.gptimageai.org/?i=d1d5k