Krea 2 Medium

Text-to-Image • krea

Smaller, faster, more cost-efficient. Extensive post-training makes outputs especially stable and consistent across generations. Strongest on illustration, anime, painting, and other expressive or artistic styles.

Model Info
Terms and License	link ↗
More information	link ↗
Pricing	View pricing in the Cloudflare dashboard ↗

const response = await env.AI.run(
  'krea/krea-2-medium',
  {
    prompt: "An igloo village glowing with Aurora's colors.",
    aspect_ratio: '1:1',
    resolution: '1K',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "krea/krea-2-medium",
  "input": {
    "prompt": "An igloo village glowing with Aurora'\''s colors.",
    "aspect_ratio": "1:1",
    "resolution": "1K"
  }
}'

Output
Raw response

{
  "state": "Completed",
  "result": {
    "image": "https://examples.aig.cloudflare.com/krea/krea-2-medium/default.png"
  },
  "gatewayMetadata": {
    "keySource": "Unified"
  }
}

Parameters

Input
Output

aspect_ratio

stringrequiredenum: 1:1, 4:3, 3:2, 16:9, 2.35:1, 4:5, 2:3, 9:16Aspect ratio of the generated image.

complexity

integerdefault: 0maximum: 100minimum: -100K2 Complexity slider (-100 to 100). 0 disables the slider LoRA.

creativity

stringdefault: lowenum: raw, low, medium, highPrompt expansion mode. `raw` disables expansion; `low`, `medium`, `high` control strength. Does not affect the K2 Intensity, Complexity, or Movement slider LoRAs.

▶image_style_references[]

arraymaxItems: 10Reference images to drive the visual style (up to 10).

intensity

integerdefault: 0maximum: 100minimum: -100K2 Intensity slider (-100 to 100). 0 disables the slider LoRA.

▶moodboards[]

arraymaxItems: 1Moodboard references (currently limited to one).

movement

integerdefault: 0maximum: 100minimum: -100K2 Movement slider (-100 to 100). 0 disables the slider LoRA.

prompt

stringrequiredText prompt describing the image to generate.

resolution

stringrequiredenum: 1KResolution scale.

seed

number | nullRandom seed for reproducible generations. Pass null or omit for a random seed.

▶styles[]

arrayStyles (typically LoRAs) to apply to the generation.

image

stringformat: uriPresigned URL for the generated image.

API Schemas (Raw)

Input

Output