谷歌推出Nano Banana Pro：全新Gemini 3专业图像生成模型
Build with Nano Banana Pro, our Gemini 3 Pro Image model
谷歌发布Gemini 3专业图像模型，可生成高保真图像，具备精准的文字渲染能力，并可通过谷歌搜索进行知识检索与实时内容对齐，在多项图像生成基准测试中领先同类竞品。
Gemini 3 Pro Image (Nano Banana Pro) is a state-of-the-art image generation and editing model. Rolls out in paid preview via Gemini API in Google AI Studio and Vertex AI for enterprises. Key capabilities: high-fidelity images, higher accuracy in text rendering, robust world knowledge. Grounding with Google Search connects the model to real-time web content for data-driven outputs. Valuable for applications requiring precise representations such as biological diagrams or historical maps. Excels on Text to Image AI benchmarks compared to leading competitors.
对 AI 行业的影响
Gemini 3图像模型的核心差异化在于将生成能力与实时知识检索结合，意味着AI生成的图像不再凭空杜撰，而是可引用真实世界数据。这对企业级应用意义重大——生物医学图表、历史地图等高精度内容生产将被大幅提速。图像生成战场正在从’以假乱真’升级为’以真驭真’。
原文参考
来源：Google DeepMind Blog · 2025-11-20