Skip to main content
POST
Gemini 3.1 Flash Image Editing
Edit images with natural language prompts using the Gemini 3.1 Flash model. Supports URL and Base64-encoded images as input, with up to 14 reference images.

Request Headers

string
required
Enum value: application/json
string
required
Bearer authentication format: Bearer {{API key}}.

Request Body

string
default:"1K"
The pixel size of the output image. 0.5K (about 512px), 1K (about 1024px), 2K (about 2048px), 4K (about 4096px). Defaults to 1K.Available values: 0.5K, 1K, 2K, 4K
object
Google Search options
string
required
Text prompt describing the desired image editing effect
string[]
List of input image URLs to edit. Supports up to 14 reference images.Array length: 0 - 14
string
The aspect ratio of the output image. Supports standard ratios and newly added ultra-wide/ultra-tall ratios (1:4, 4:1, 1:8, 8:1)Available values: 1:1, 1:4, 1:8, 2:3, 3:2, 3:4, 4:1, 4:3, 4:5, 5:4, 8:1, 9:16, 16:9, 21:9
string[]
List of Base64-encoded images to edit. The total number of image_urls and image_base64s cannot exceed 14.Array length: 0 - 14
string
default:"image/png"
The MIME type of the output image. Supported formats: image/png, image/jpeg. Defaults to image/png.Available values: image/png, image/jpeg

Response Information

string[]
required
List of edited image URLs
object
Model grounding metadata. Returned when Google server-side tools (such as Search) are called. Note: Tool calls depend on Google. Enabling tools does not guarantee that a tool call will be triggered 100% of the time.