Nano Banana 2 Lite

The fastest, most cost-efficient image generation model in the Nano Banana family

Context window

Not announced

Input / 1M tokens

Not announced

Output / 1M tokens

33.6

Provider

Google

Data verified 2026-07-04

Nano Banana 2 Lite is Google's entry-level image generation model optimized for speed and cost efficiency. It generates text-to-image outputs in approximately 4 seconds at $0.034 per 1,000 images, positioning it as the lowest-cost option in Google's image model family. Designed for high-throughput, rapid iteration workflows where speed and cost matter more than advanced reasoning.

Capability index

Relative estimates (0-100) to place this model against its peers, grounded in published benchmarks.

Coding

Reasoning

Math

Multimodal

Long context

Speed

Cost efficiency

How to access it

Available through Google AI Studio, Gemini API, and Gemini Enterprise Agent Platform. Rolling out to consumer surfaces including AI Mode in Search, Gemini app, NotebookLM, Google Photos, Stitch, Google Flow, and Google Ads.

Get access →Documentation →

Strengths

✓Ultra-fast 4-second generation time
✓Lowest cost in Google's image lineup
✓High throughput for batch generation
✓Strong character consistency
✓Legible in-image text rendering

Best for developers who...

High-volume image generationRapid ideation and A/B testingIterative content creationSocial media and marketing workflowsReal-time creative applications

When to choose it (and when not to)

Reach for Nano Banana 2 Lite when...

→When cost-per-sample is the primary constraint
→For workflows requiring 100+ images per session
→When generation latency must be under 5 seconds
→For rapid prototyping and exploration

Look elsewhere if...

✕Use cases requiring advanced reasoning or complex scene composition
✕When legible signage, infographic data, or narrative continuity is critical
✕Projects where maximum fidelity and detail are essential
✕Applications needing search grounding for real-world context

How to use it

›Use concise, direct prompts for optimal results
›Include style references in descriptions
›Specify desired aspect ratios explicitly
›Chain with Gemini Omni Flash for image-to-video workflows

Quickstart

Python

import google.generativeai as genai
genai.configure(api_key='YOUR_API_KEY')
model = genai.ImageGenerationModel('gemini-3.1-flash-lite-image')
response = model.generate_images(prompt='A sunset over mountains')

Visit Google AI Studio for playground access without API key

API model id: gemini-3.1-flash-lite-image

Benchmarks

Benchmark	Score	Notes
Arena.ai Text-to-Image Leaderboard	Ranked #5	Positioned almost as highly as non-Lite Nano Banana 2 despite lower latency
Text-to-Image Elo Score	1251	Exceeds Nano Banana Pro (1245)
Single-Image Editing Elo Score	1308
Multiple-Image Editing Elo Score	1294

Source: Arena.ai

Learn the concepts

Inference Context Window

← All AI models