Connect

Image generation

Last updated: November 6, 2026

CoreCode generates images through the OpenAI Images API: the GPT group has gpt-image-2, gpt-image-2.5-flare and gpt-image-2.5-sunburst, and the Grok group has grok-imagine-image. Call them with the OpenAI SDK or curl; each image is billed at a fixed price.

Models

To get a key, click "Create API key" on the API Keys page and pick the group from the table.

Generate with curl

curl https://corecode.live/v1/images/generations \
  -H "Authorization: Bearer sk-your-key" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-image-2.5-flare", "prompt": "a small red fox sitting on a rock, flat illustration", "quality": "low"}' \
  | jq -r '.data[0].b64_json' | base64 --decode > fox.png

The image comes back as base64 in data[0].b64_json. To switch models, change model.

Generate with the Python SDK

import base64
from openai import OpenAI

client = OpenAI(base_url="https://corecode.live/v1", api_key="sk-your-key", timeout=180)
r = client.images.generate(
    model="gpt-image-2.5-flare",
    prompt="a small red fox sitting on a rock, flat illustration",
    quality="low",
)
open("fox.png", "wb").write(base64.b64decode(r.data[0].b64_json))

An image takes over ten seconds, so set the client timeout to 60 seconds or more.

Edit an image

gpt-image-2, gpt-image-2.5-flare and gpt-image-2.5-sunburst can edit: as with OpenAI, upload the source image as the form field image.

curl https://corecode.live/v1/images/edits \
  -H "Authorization: Bearer sk-your-key" \
  -F model=gpt-image-2.5-flare \
  -F "prompt=give the fox a blue scarf" \
  -F quality=low \
  -F image=@fox.png \
  | jq -r '.data[0].b64_json' | base64 --decode > fox-scarf.png

In the SDK: client.images.edit(model=..., image=open("fox.png", "rb"), prompt=...).

Generate with Grok

curl https://corecode.live/v1/images/generations \
  -H "Authorization: Bearer sk-your-key" \
  -H "Content-Type: application/json" \
  -d '{"model": "grok-imagine-image", "prompt": "a small red fox in the snow, watercolor", "response_format": "b64_json"}' \
  | jq -r '.data[0].b64_json' | base64 --decode > fox.jpg

Pass "response_format": "b64_json". Without it you get data[0].url, a temporary xAI link that may not open from mainland China.

Parameters

FAQ