Images

GPT Image

OpenAI’s GPT Image models: sharp detail, strong text rendering, and editing with up to 16 images. GPT Image 2 costs one price at every size; the 2.6 and 2.5 models, Flare and Sunburst, are priced by size.

The models

ModelIDPrice per imageReference images
GPT Image 2.6 Flaregpt-image-2.6-flare$0.008up to 16
GPT Image 2.6 Sunburstgpt-image-2.6-sunburst$0.008up to 16
GPT Image 2.5 Flaregpt-image-2.5-flare$0.008up to 16
GPT Image 2.5 Sunburstgpt-image-2.5-sunburst$0.008up to 16
GPT Image 2gpt-image-2$0.02up to 16

GPT Image 2.6 is the newest, from November 2026: Flare for fast iteration, Sunburst for detailed visual work, at the same price. GPT Image 2.5 Flare and Sunburst are their September versions, and GPT Image 2 stays for workflows tuned to it. GPT Image 2.6 also makes animated GIFs (below). OpenAI’s SVG output, complexity setting and markup edits aren’t available on Zurelay yet.

Make an image

import os
from openai import OpenAI
client = OpenAI(base_url="https://api.zurelay.com/v1", api_key=os.environ["ZURELAY_API_KEY"])
result = client.images.generate(
model="gpt-image-2.6-sunburst",
prompt="An isometric illustration of a tiny bakery at night, warm window light, a cat on the awning",
size="1536x1024",
extra_body={
"quality": "high"
},
)
print(result.data[0].url)

Parameters

sizestring
1024x1024, 1536x1024, 1024x1536, 2048x2048, 3840x2160, 2160x3840, or auto. GPT Image 2 costs the same at every size. Flare and Sunburst price up to 1024x1024 as 1K, up to 2560x1440 as 2K, and larger as 4K; with no size or auto they make 1024x1024.
qualitystring
low, medium or high (the default). Same price at every level.
backgroundstring
transparent, opaque or auto. Transparent needs PNG or WebP output.
output_formatstring
png, jpeg or webp. GPT Image 2.6 usually answers in PNG whatever you ask. On 2.6, gif makes an animated GIF (below).
ninteger
1 to 4 images.

Animated GIFs

GPT Image 2.6 Flare and Sunburst make animated GIFs: set output_format to gif on /v1/images/generations. Describe the motion in the prompt, and say it should loop if it should. A GIF has 9 frames; size is each frame’s.

import os
from openai import OpenAI
client = OpenAI(base_url="https://api.zurelay.com/v1", api_key=os.environ["ZURELAY_API_KEY"])
result = client.images.generate(
model="gpt-image-2.6-flare",
prompt="A cheerful cartoon fox waving hello, flat illustration, simple background, seamless loop",
size="768x768",
extra_body={
"output_format": "gif",
"frame_duration_ms": 150
},
)
print(result.data[0].url)
sizestring
Each frame’s size, WIDTHxHEIGHT in multiples of 16 from about 272x272 to 960x960, at a ratio from 1:3 to 3:1 (768x432, 896x512). Default 512x512.
frame_duration_msinteger
How long each frame shows, 20 to 2000. Default 200.
frame_countinteger
Always 9 for now; leave it out.
  • Pricing: $0.028 with frames up to 512 × 512, $0.045 up to 768 × 768 and $0.064 up to 960 × 960, the same on Flare and Sunburst.
  • GIFs are made at OpenAI’s low GIF quality, so a GIF’s price depends on its frame size alone; quality is ignored.
  • GIFs come from a prompt alone: no reference images, and /v1/images/edits can’t make them.
  • Most arrive in 15 to 35 seconds.

Edit with up to 16 images

Use client.images.edit with one or more files, or JSON with images. Masks aren’t supported; describe the area to change.

Python
result = client.images.edit(
model="gpt-image-2.6-flare",
image=[open(f"tile-{i}.png", "rb") for i in range(1, 9)],
prompt="Arrange the eight product photos into a clean 4 x 2 catalog grid on white.",
)

More on sending images: Image editing and reference images.

Questions, or something missing? Ask our support team and we’ll answer by email, usually within a minute.