gemini-3.1-flash-lite-image with the Gemini generateContent request format at api.yingtu.ai.
The model is a strong fit when every delivered asset can remain at 1K. Choose Gemini 3.1 Flash Image instead when a product needs 2K, 4K, or extended aspect ratios.
The
-image suffix is required. gemini-3.1-flash-lite-preview refers to a similarly named text model and cannot generate images.Gemini 3.1 Flash Lite Image API at a glance
Google describes the upstream model as a low-latency, high-volume image model. Review the Gemini 3.1 Flash Lite Image model page for the upstream positioning; provider latency targets are not an end-to-end gateway SLA.
When to use Gemini 3.1 Flash Lite Image
High-volume 1K assets
Lite works well for product thumbnails, in-app cards, editorial illustrations, social media assets, and other workflows where additional resolution would not change the delivered result. A 1K-only pipeline also keeps response payloads and storage requirements predictable.Interactive image editing
Place an edit instruction and one or more Base64 source images in the sameparts array. This supports background replacement, style changes, reference combination, and composition updates.
Migration from Gemini 2.5 Flash Image
Teams leavinggemini-2.5-flash-image can evaluate Lite first when the existing workflow already targets 1K. Migration testing should compare prompt adherence, edit behavior, returned MIME types, error handling, and end-to-end latency—not only the model ID.
Limits that affect an integration
imageSize may be omitted or set to uppercase 1K. Values such as 2K, 4K, 1k, or 1024px are outside this model’s public contract.
The supported aspect ratios are 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, and 21:9. See the parameters reference for the complete request structure.
Images are returned in candidates[].content.parts[].inlineData. Scan every part and use the returned mimeType when choosing a file extension. Do not assume the image is always parts[0] or always PNG.
Lite versus Gemini 3.1 Flash Image
Use the dated pricing page for exact rates. If higher resolutions are likely to enter the roadmap, starting with
gemini-3.1-flash-image avoids a later model migration.
Gateway request capacity, upstream 429 behavior, and the availability objective are documented separately under traffic capacity and availability. They do not turn Google’s low-latency positioning into a per-image completion guarantee.
Call the API
Send a synchronous request toPOST /v1beta/models/gemini-3.1-flash-lite-image:generateContent. For work that must outlive the client connection, use /v1/images/tasks with action=generate|edit, store task_id, and retrieve GET /v1/tasks/{task_id} until a terminal state.
Configure Lite parameters
Set the 1K output, aspect ratio, response modalities, and source image.
Run Lite examples
Copy synchronous, asynchronous, editing, and response-parsing code.