GPT Image 1.5 I2I: AI Image Editor Online
Bottom line
OpenAI's instruction-driven image editor — up to 10 inputs, medium or high quality, strict content filters. Open the Image Edit tool, select GPT Image 1.5 I2I, upload your image, and the exact credit cost appears before the edit runs.
The argument for GPT Image 1.5 I2I is comprehension. Because it comes from OpenAI's multimodal GPT stack, it parses editing instructions the way a language model parses text — long, conditional, multi-step prompts such as "swap the jacket to denim, keep the logo intact, match the lighting from the left" tend to survive as written instead of collapsing into an approximate restyle.
The spec sheet is deliberately narrow. You can attach up to 10 images per edit, choose between medium and high quality, and work in three aspect ratios: square 1:1, portrait 2:3, or landscape 3:2. There is no ultrawide format here; the model spends its effort on executing the requested change faithfully at sensible sizes.
Two things to know before you commit. It carries VdoBloom's CENSORED label — the strictest content tier, reflecting OpenAI's conservative provider filters — and it holds a 4.4 user rating. Jobs are billed in VdoBloom credits, with the cost displayed before every run.
What creators use GPT Image 1.5 I2I for
- Multi-step product edits where instructions must be followed literally — change the color, preserve the label text, adjust the shadows.
- Instruction-driven composites that pull from up to 10 reference photos in one job.
- Edits involving legible text — packaging, signage, UI mockups — where prompt comprehension matters more than raw resolution.
How to use GPT Image 1.5 I2I on VdoBloom
- 1Open the Image Edit tool and select GPT Image 1.5 I2I.
- 2Upload the image(s) to edit and describe the change you want.
- 3Review the credit cost and generate — results appear in your library.
Frequently asked questions
How many images can I attach to a GPT Image 1.5 I2I edit?
Up to 10 per job. The model treats them as joint context for your instruction, which is useful when a single change needs to reference several sources — a subject from one photo, a texture from another, a layout from a third.
What do the medium and high quality settings change?
They control how much rendering effort goes into the output: high produces cleaner detail, while medium is the quicker, lighter option for drafts. VdoBloom shows the credit cost for whichever setting you pick before you generate.
Will GPT Image 1.5 I2I refuse my prompt?
It can — the model carries the CENSORED label, VdoBloom's strictest content tier, reflecting OpenAI's conservative provider filters. Requests involving real people, violence, or suggestive content have a higher chance of being declined than on MODERATE or HOT models.
Why does GPT Image 1.5 I2I only offer 1:1, 2:3, and 3:2 aspect ratios?
Those are the frames the model natively supports. If you need ultrawide or other formats, generate at 3:2 or 2:3 and re-frame afterward, or choose a different editor on VdoBloom with a wider ratio list.