ComfyUI Extension: ComfyUI-YK-Banana2-Batch

Authored by Bzbaozi

Created

Updated

1 stars

A powerful batch image generation node for ComfyUI using the YK-Banana2 Pro API to generate multiple images in parallel with reference image support. (Description by CC)

Custom Nodes (0)

    README

    🍌 ComfyUI-YK-Banana2-Batch

    🍌 ComfyUI-YK-香蕉2 批量生成节点

    A powerful batch image generation node for ComfyUI that leverages the YK-Banana2 Pro API to generate multiple images in parallel with reference image support.
    一个强大的 批量图像生成节点,基于 YK-香蕉2 Pro API,支持参考图并行生成多张图像。

    💡 Perfect for fashion transfer, character customization, and multi-prompt batch workflows.
    💡 适用于服装迁移、角色换装、多提示词批量生成等场景。


    ✨ Features / 功能特点

    • Batch processing: Run up to 10 independent tasks simultaneously
      批量处理:最多同时运行 10 个独立任务
    • 🖼️ Reference image support: Each task accepts up to 3 reference images
      🖼️ 参考图支持:每个任务可输入最多 3 张参考图
    • 📝 Per-task prompts: Customize prompt for each batch item
      📝 独立提示词:每个任务可设置专属文本提示
    • 🔍 Preserve original resolution: No forced resize or interpolation
      🔍 保留原始分辨率:不强制缩放或插值
    • ⏱️ Parallel execution: Faster than sequential generation
      ⏱️ 并行执行:比逐个生成更快
    • 🧩 Seamless ComfyUI integration: Outputs standard IMAGE tensors
      🧩 无缝集成 ComfyUI:输出标准 IMAGE 张量

    🔑 Requirements / 所需条件

    • A valid API key from YK-AI
      👉 Contact on WeChat: v:YK-ai001
    • 有效的 YK-AI API 密钥
      👉 微信联系:v:YK-ai001

    ⚠️ This node requires internet access and an active YK-Banana2 Pro subscription.
    ⚠️ 本节点需要联网,并依赖有效的 YK-香蕉2 Pro 订阅服务。


    📥 Installation / 安装方法

    1. Open your ComfyUI directory
      进入你的 ComfyUI 根目录
    2. Navigate to the custom_nodes folder
      进入 custom_nodes 文件夹
    3. Clone this repository:
      克隆本仓库:
    cd custom_nodes
    git clone https://github.com/YOUR_GITHUB_USERNAME/ComfyUI-YK-Banana2-Batch.git
    Restart ComfyUI
    重启 ComfyUI
    The node will appear as YK_Banana2_Batch in the node menu.
    
    节点将在菜单中显示为 YK_Banana2_Batch。
    
    🧪 Usage Guide / 使用说明
    Input Slots (per task i = 1..10)
    输入接口(每个任务 i = 1..10)
    Input / 输入	Type / 类型	Description / 说明
    image_i_1	IMAGE	Reference image 1 (e.g., person) / 参考图1(如人物)
    image_i_2	IMAGE	Reference image 2 (e.g., clothing) / 参考图2(如服装)
    image_i_3	IMAGE	Optional third reference / 可选第三张参考图
    prompt_i	STRING	Prompt for this task / 本任务的提示词
    📌 Example prompt / 示例提示词:
    
    “将图2的衣服穿在图1人物的身上,保持图1的人物不变,保持图1的场景不变”
    
    "Wear the clothes from image 2 on the person in image 1, keep the person, scene, and pose from image 1 unchanged."
    
    Output Slots / 输出接口
    output_1 to output_10: Generated images (IMAGE type)
    output_1 到 output_10:生成的图像(IMAGE 类型)
    Unused outputs return blank images (safe to ignore)
    未使用的输出返回空白图(可安全忽略)
    🛠️ Configuration / 配置方式
    Set your API key via environment variable (recommended):
    
    通过环境变量设置 API 密钥(推荐):
    
    Bash
    编辑
    export YK_BANANA2_API_KEY="your_api_key_here"
    Or enter it directly in the node’s api_key field (less secure).
    
    或直接在节点的 api_key 输入框中填写(安全性较低)。
    
    📞 Support & Contact / 支持与联系
    For API access, bug reports, or feature requests:
    
    如需获取 API、反馈 Bug 或提出功能建议:
    
    💬 WeChat / 微信: v:YK-ai001
    
    📜 License / 开源协议
    This project is licensed under the MIT License – see LICENSE for details.
    
    本项目采用 MIT 许可证,详情见 LICENSE 文件。