Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)
Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) is the fastest, most cost-efficient image model in the Nano Banana family, generating and editing 1K images with reliable prompt adherence, strong character consistency, and legible in-image text for high-volume production workloads.
import { generateText } from 'ai'
const result = await generateText({ model: 'google/gemini-3.1-flash-lite-image', prompt: 'Render a picture of a red balloon.',});About Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)
Google released Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) on June 30, 2026 as the fastest, most cost-efficient image model in the Nano Banana family. Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) is built for high-throughput pipelines where teams generate thousands of images and iterate quickly, keeping per-image cost low enough for near-real-time and high-volume workflows.
Despite the speed focus, Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) retains the traits the Nano Banana family is known for: reliable prompt adherence, strong character consistency, and legible in-image text rendering. Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) handles text-to-image generation, prompt-based editing, and multi-image composition through a single multimodal interface, so a generation pipeline and an editing pipeline can share one model.
The main tradeoff is resolution. Output is capped at 1K, while other models in the family generate at 2K and 4K. Google positions the lineup in three tiers: Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) for speed and scale, Gemini 3.1 Flash Image as the generalist, and Gemini 3 Pro Image for complex professional work. Google also recommends Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) as the upgrade path from the original Nano Banana, google/gemini-2.5-flash-image, citing better quality, faster generation, and lower cost.
Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) accepts prompts within a context window of 65.5K tokens and returns text and images in the same response. Every generated image carries an invisible SynthID watermark for downstream identification of AI-generated content.