Generate image

Read and creategenerate_image

Generates one image from a prompt in your Brand Kit's style and saves it to the Media Library. Give draftId to put it straight onto a draft. Uses AI credits.

Needs the write scope. Creates or changes things in Chirpy. Publishes nothing.

Inputs

Input	Type	Required	Description
`prompt`	string	Yes	Subject, composition, background, mood, lighting, style; leave clear space if text will be added later
`aspectRatio`	"1:1" \| "4:5" \| "9:16" \| "16:9"
`useBrandStyle`	boolean		Apply the Brand Kit palette and logo (default true)
`draftId`	integer		A draft to add the image to once it is made
`referenceFileIds`	integer[]		Up to 5 Media Library image fileIds to use as references: the user's photos, product or style to follow. Use the ids of images the user attached in this chat, or ones you found with list_media and look_at_media. Up to 5.
`referenceMode`	"inspiration" \| "variant"		inspiration (default): the references guide the look; variant: keep the reference's subject, composition and palette and refresh the rest

REST

curl https://chirpy.me/api/v1/generate_image \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{"prompt":"..."}'

MCP

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "generate_image",
    "arguments": {
      "prompt": "..."
    }
  }
}

MCP hints: readOnlyHint: false, destructiveHint: false, idempotentHint: false, openWorldHint: false.