Direct the product shot
Describe the finish, camera angle and lighting. A clear reference helps communicate the shape and details that matter to your brand.
OpenAI Image generation & editing
Create and edit images with GPT Image 2.5 Sunburst in Riverflow, OpenAI’s variant focused on precise changes and reference fidelity.
Open Riverflow
The example
Amber glass, lemon peel and morning light.
Create a premium fragrance campaign photograph, landscape 4:3. A low rectangular bottle of translucent amber perfume with a brushed aluminium cap stands on a cream travertine shelf. The bottle has a beautifully typeset small ivory paper label reading exactly AFTER RAIN on two lines, with EAU DE PARFUM underneath. A long curl of fresh lemon peel rests beside it, and one wet dark green leaf touches the edge of the shelf. Low morning sunlight passes through the bottle and casts a luminous amber rectangle and crisp shadows onto the stone. Behind it is a pale blue-grey plaster wall, clean and quiet. Art direction: tactile, architectural, editorial, close-up 70mm still-life photography, subtle natural grain, tiny droplets on the glass. Place the bottle in the right half of the image, leaving elegant unbroken space on the left. Every object rests on the shelf; no levitation, no decorative graphics, no watermark, no extra words. The complete bottle and cap must fit in the frame.
Select GPT Image 2.5 after opening the image tool.
GPT Image 2.5 is an image model from OpenAI for text-to-image generation and multi-reference image editing. It comes in different variants. Riverflow supports Sunburst, the variant focused on higher-precision editing. It is better suited to precise changes, with less emphasis on creative exploration and divergent results.
Choose GPT Image 2.5 in Riverflow’s image tool, add your brief and references, then review the settings and credit cost before generating.
Available settings depend on the model and workflow.
Plans and credits| Model | 1K | 2K | 4K |
|---|---|---|---|
| GPT Image 2.5 High | 18 credits | 18 credits | 40 credits |
One image uses the amount shown. Four 2K images use four times the 2K cost. Image edits can use a different rate; check the quote in the editor. Your complete quote appears before generation.
Describe the finish, camera angle and lighting. A clear reference helps communicate the shape and details that matter to your brand.
Build a composition around the space you need for copy. Choose the output shape in Riverflow and describe where the subject should sit.
Request a transparent background in an image edit for use in another design. Review the cutout around glass, hair and fine edges.
Models, inputs and getting started.
GPT Image 2.5 is an image model from OpenAI for text-to-image generation and multi-reference image editing. It comes in different variants. Riverflow supports Sunburst, the variant focused on higher-precision editing. It is better suited to precise changes, with less emphasis on creative exploration and divergent results.
The image-editing workflow supports native transparency for GPT Image 2.5. Request transparency explicitly and use a supported output format. Check the finished result, especially hair, glass and product edges; output quality still depends on the source and brief.
Sign in to Riverflow, open the image creation tool and choose GPT Image 2.5 from the model picker. Add your brief and any required references, then review the available settings and credit cost before generating.
GPT Image 2.5 High costs 18 credits per 1K image, 18 credits per 2K image and 40 credits per 4K image in Riverflow’s image generator. Generating multiple images multiplies that cost. Credits are included in Riverflow subscriptions; see the pricing page for plan allowances.
Bring your brief. Choose your model. Start creating.