GPT Image 2.5

GPT Image 2.5 is OpenAI’s image series that topped the Image Arena. Two tiers: gpt-image-2.5-flare — the fast one, for everyday generation, batches and prototypes; gpt-image-2.5-sunburst — the precision tier, for accurate edits and final advertising images. Both work with POST /v1/images/generations and POST /v1/images/edits.

Parameters

Field	Description
model	openai/gpt-image-2.5-flare or openai/gpt-image-2.5-sunburst (short gpt-image-2.5-flare / gpt-image-2.5-sunburst)
prompt	What to draw, required; up to 20,000 characters — paste briefs whole
n	How many images, 1 to 4; each is billed separately
aspect_ratio	Ratios from 1:3 to 3:1, including 2:1, 1:2, 3:1, 1:3, 21:9; defaults to 1:1
resolution	1k, 2k or 4k; defaults to 1k
size	Explicit WxH (both edges multiples of 16, long side up to 3840)
background	transparent | opaque | auto — the series’ background knob
output_format	png | jpeg | webp; defaults to png
quality	The series has no such knob: passing it returns 400

Transparent background

background: "transparent" cuts the object out with no backdrop — the output format must then be png or webp (JPEG has no alpha), otherwise 400. auto lets the model decide from the prompt, opaque forces a filled background — handy when the prompt leans toward transparency but you don’t want it.

curl https://api.mixen.ai/v1/images/generations \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-image-2.5-flare",
    "prompt": "Sticker: capybara astronaut, outline style",
    "aspect_ratio": "1:1",
    "background": "transparent",
    "output_format": "png"
  }'

Editing from references — up to 16 images

/v1/images/edits accepts up to 16 references per request — combine products, style references and characters:

curl https://api.mixen.ai/v1/images/edits \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -F model=openai/gpt-image-2.5-sunburst \
  -F image=@model.png \
  -F image=@jacket.png \
  -F image=@studio-light.png \
  -F prompt="Dress the model in this jacket, lighting as in the third reference"

Text-to-image

Python

import base64

img = client.images.generate(
    model="openai/gpt-image-2.5-sunburst",
    prompt="Final advertising shot: perfume on a marble counter, water drops, "
           "side lighting, photorealistic",
    extra_body={"aspect_ratio": "3:1", "resolution": "4k"},
)
with open("banner.png", "wb") as f:
    f.write(base64.b64decode(img.data[0].b64_json))

cURL

curl https://api.mixen.ai/v1/images/generations \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-image-2.5-sunburst",
    "prompt": "Final advertising shot: perfume on a marble counter, water drops, side lighting, photorealistic",
    "aspect_ratio": "3:1",
    "resolution": "4k"
  }'

Pricing

$0.045 per image for both tiers — the flat catalog price. Current values — catalog.

Strengths and limitations

• Two tiers per task: Flare — speed and batches, Sunburst — precision edits.
• Native transparent background (background + png/webp).
• Up to 16 references on /v1/images/edits; prompts up to 20,000 characters.
• Aspect ratios from 1:3 to 3:1, sizes up to 4K (3840x2160; above 2560x1440 the vendor marks them experimental).
• No quality knob; background/output_format sent to other models return 400.
• Balance-only — not part of the free quota.

All models — in the catalog. General image workflow — images guide.