AI 模型 · google

Google Nano Banana 2

Fast image generation with conversational edits, fusion, and consistent characters.

10 CR / RUN · google · 11.2M RUNS

简介

nano-banana-2 is Google’s fast image generation model for creating and refining visuals with natural-language prompts. It supports straightforward text-to-image workflows while also handling image-guided generation through image_input.

What makes it especially flexible is its support for conversational editing, multi-image fusion, and character consistency. With controls for resolution, aspect ratio, output format, and optional image or Google search inputs, it’s built for iterative creative work as well as quick visual production.

优势

01Fast image generation from text prompts
02Conversational editing for iterative refinement
03Multi-image fusion for blended concepts
04Character consistency across generations
05Flexible controls for resolution and aspect ratio
06Supports image-guided creative workflows

示例提示词

Create a cinematic portrait of a futuristic street racer at dusk, neon reflections on wet pavement, dramatic lighting, highly detailed.

试试这个提示词

Using the reference image, turn this living room into a warm Scandinavian interior with light oak furniture, soft textiles, and natural daylight.

试试这个提示词

Blend a vintage botanical illustration style with a modern skincare product shot, clean background, elegant composition, premium branding feel.

试试这个提示词

Design a consistent character sheet for a cheerful young wizard, showing front view, side view, and three expressive poses in a colorful storybook style.

试试这个提示词

使用场景

Rapid concept generation

Generate polished visual ideas quickly from a simple text prompt. It’s a strong fit for early-stage ideation, moodboards, and creative exploration.

Iterative image editing

Refine images through conversational edits instead of restarting from scratch. This makes it easier to adjust style, composition, or details over multiple passes.