Google Unveils Nano Banana Pro With Major Image-Gen Upgrades
谷歌发布 Nano Banana Pro，大幅提升图像生成能力
Google has launched Nano Banana Pro, an advanced image-generation model built on its new Gemini 3 LLM. The model delivers sharper images, improved text rendering, and detailed stylistic control, including lighting, camera angles, depth of field, and color grading. It also adds web-search abilities for tasks like finding recipes or creating flash cards. Resolution now reaches 2K and 4K, compared with the original model’s 1024×1024 limit. However, the upgrade comes at higher costs: $0.139 per 1080p/2K image and $0.24 for 4K. The system supports up to six high-fidelity shots, blends 14 objects, and keeps consistency across five people. Nano Banana Pro is rolling out across Gemini, NotebookLM, Workspace apps, and Google’s developer tools. Google has also integrated SynthID for watermarking and plans broader C2PA support.
谷歌推出基于最新 Gemini 3 模型的 Nano Banana Pro，带来更清晰图像、更准确文字渲染和更精细的风格控制，包括光线、镜头角度、景深与调色等。新模型还具备网页搜索能力，可查找食谱或生成学习卡片。分辨率提高到 2K、4K，但生成成本上升：1080p/2K 每张 0.139 美元，4K 每张 0.24 美元。系统支持六张高保真输入、融合 14 个物体，并能保持五个人物的一致性。Nano Banana Pro 已整合至 Gemini、NotebookLM、Workspace 及开发者工具。谷歌同步加入 SynthID 水印技术，并计划进一步支持 C2PA 标准。
