GPT Image 2.5
OpenAI · AI imageGPT Image 2.5 Sunburst is OpenAI's most capable image model for text-to-image and precise image editing. It follows instructions closely while preserving identity, composition, and lighting on the parts you want unchanged.
On VidMachine it is 10 credits per image. Native aspect ratios include 9:16 and 16:9.
Key features and benefits
Generation and precise editing
Generate from a prompt alone, or pass reference images to edit or combine styles and subjects. Sunburst is tuned for editing workflows where fidelity to the input matters.
Photorealism and text
Strong materials, lighting, and readable on-image text when you quote exact copy and describe typography—useful for thumbnails, UI mockups, and marketing frames.
Native vertical and landscape ratios
Supports 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, and 9:16, including the vertical and landscape frames used for shorts.
Predictable cost
10 credits per image on VidMachine, for generation and edits.
Technical specifications
ProviderOpenAI
ModesText-to-image; image editing with references
Aspect ratios1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16
Credits10 per image
Use cases and applications
Use GPT Image 2.5 when you want OpenAI's latest stills with strong instruction following and native 9:16 for Shorts.
Prefer it for targeted edits—style transfer, object changes, or multi-reference composites—where preserving the original composition matters.
Why this model
Choose GPT Image 2.5 Sunburst when you want the newer OpenAI image stack with native vertical ratios and editing precision.
What you should know
Does it support vertical frames?
Yes. 9:16 and 16:9 are supported, along with square and other common ratios.
How much does GPT Image 2.5 cost?
10 credits per generated image on VidMachine.