GPT Image 2.5 Sunburst Edit
OpenAI GPT Image 2.5 Sunburst Edit, the precision tier of the 2.5 edit endpoint. Slower than Flare Edit, better at holding faces, fine product geometry and small label type. Same price as Flare Edit. Edit once from the clean source rather than re-editing an edit (each pass adds noise and drift); xhigh and max cost more credits than high (max is 2 to 4x), so keep them for final frames. Sizes up to 2560x1440 are reliable; larger is experimental.
Try GPT Image 2.5 Sunburst Edit in the playground
GPT Image 2.5 Sunburst Edit
Node type: edit:openai_gpt_image_2_5_sunburst_edit
Category: edit
Description
GPT Image 2.5 Sunburst Edit is the precision tier of OpenAI's GPT Image 2.5 editing: wire the source image into Image 1, add an optional Mask, describe the change in Prompt, and run the node for 1 to 4 edited images. It is slower than Flare Edit but better at holding faces, fine product geometry and small label type, which makes it the place for final frames. Edit once from the clean source, because each pass over an earlier edit adds noise and drift. Sizes up to 2560x1440 are reliable and larger ones are experimental; inputs wider or taller than 3:1 need cropping or a custom output size. It costs about 20 credits per image, and the xhigh and max quality settings cost more than high (max is 2 to 4x).
Pricing
- Cost: ~20 credits per images
Variant family
- Family:
gpt-image-2.5-sunburst - Is primary: no
- All variants:
generate:openai_gpt_image_2_5_sunburst_text_to_image,edit:openai_gpt_image_2_5_sunburst_edit
Canvas ports
These appear as port handles on the left side of the node.
| ID | Label | Details |
|---|---|---|
prompt |
Prompt | TEXT (required) · multiline |
image1 |
Image 1 | IMAGE (required) |
mask_url |
Mask | IMAGE |
Sidebar config
These render as form fields in the right-side config panel when the node is selected.
| ID | Label | Details |
|---|---|---|
quality |
Quality | TEXT · options: low, medium, high, xhigh, max |
image_size |
Image Size | IMAGE_SIZE · options: custom, square_hd, portrait_4_3, landscape_4_3, auto, +1 — Match input keeps the first image’s resolution, rounded to 16px. Sizes are scaled to fit 0.66 to 8.29 megapixels and a 3840px edge. Inputs wider or taller than 3:1 need cropping or a custom output size. Custom sizes above 2560x1440 are experimental and can come out over-sharpened. |
num_images |
Number of Images | NUMBER · 1…4 |
output_format |
Output Format | TEXT · options: jpeg, png, webp |
background |
Background | TEXT · options: auto, transparent, opaque |
output_compression |
Output Compression | NUMBER · 0…100 |
sync_mode |
Sync Mode | BOOLEAN |
Outputs
| ID | Label | Type |
|---|---|---|
images |
Generated Images | IMAGE |
Realistic avatar stills
Building a talking-head avatar? Wiring any image into the Sunburst generate node routes here, with the first wired photo landing on image1. Make a Realistic AI Avatar has the rule: keep image1 a face-free scene plate, put identity refs in image2 and up, labelled "identity only," and describe the shot in the prompt. Also covers the masked-edit fix for one bad panel on a sheet.
Auto-generated from the Wireflow node registry.