ImgsGen / MODELS

Gemini AI Image Generator: Models, Sizes and Workflow

Describe a scene in English or Chinese, select a target resolution and composition, and explore the resulting image.

By ImgsGen · Updated

At a glance

Gemini Image offers target resolution and aspect-ratio controls in ImgsGen, with automatic quality. Multiple images are generated sequentially. Reference-image editing is not currently enabled.

Available models

The studio catalog includes Gemini 3 Pro Image, Gemini 3.1 Flash Image and Gemini Nano Banana 2.1. Use a shared prompt when exploring different interpretations. Capabilities are defined by the controls available in this studio rather than by assuming every named model has identical features.

Resolution and aspect ratio

Choose 1K, 2K or 4K target tiers and a supported aspect ratio. Quality is automatic. Actual pixel dimensions may differ by model and composition, so read the result dimensions and inspect the downloaded picture. The linked live example uses Gemini Flash at 1K.

Generating more than one image

A batch is processed one image at a time. Larger batches can take longer, and the studio preserves finished images if a later request fails. Check history before resubmitting an interrupted request to avoid unnecessary duplicate generations.

Continue from an existing picture

Reference editing is not currently available for Gemini in this studio. If you need the previous picture to guide the next generation, select a supported GPT Image model and attach it. Reusing the same text alone does not preserve an exact composition.

Bring the idea to your canvas.

Choose a supported model and composition. The studio uses an ImgsGen account and subscription credits.

Open image studio ↗