Package Information
Documentation
n8n-nodes-qiniu-ai
🇺🇸 English
n8n community node for Qiniu Cloud AI SDK - Integrate full-modal AI capabilities into your n8n workflows.
✨ Features
| Resource | Operations | Description |
|---|---|---|
| Chat | Complete | Multi-model chat completion (Qwen, Claude, Gemini, GPT, DeepSeek, etc.) |
| Image | Generate, Edit | AI image generation and editing with multiple models |
| Video | Generate, Remix, Get Status | Video generation with Kling, Veo, Sora models |
| Audio | Text-to-Speech, Speech-to-Text | TTS and ASR capabilities |
| Agent | Execute | AI agent with built-in tools (Web Search, OCR, Image/Video Generation) and ReAct loop |
| Tools | Web Search, OCR, Image Censor, Video Censor, VFrame | Utility tools with content safety and video processing |
🆕 What's New in v1.0.0
- Cloud-Native State Persistence (KodoCheckpointer): Store agent conversation state in Qiniu Kodo object storage for production-grade persistence across workflow executions.
- Content Safety Tools:
- Image Censor: Synchronous content safety audit (pulp, terror, politician detection)
- Video Censor: Asynchronous video content moderation with job polling
- Video Frame Extract (VFrame): Extract frames from video at specific timestamps
- SDK v0.27.3: Upgraded to the latest SDK with enhanced Agent capabilities.
📦 Installation
Community Node (Recommended)
- Go to Settings > Community Nodes
- Click Install
- Enter
n8n-nodes-qiniu-aiand click Install
Manual Installation
# In your n8n custom nodes directory
npm install n8n-nodes-qiniu-ai
🔧 Configuration
Create credentials in n8n:
- Go to Credentials > New
- Search for Qiniu AI API
- Enter your API Key (obtain from Qiniu Cloud Console)
(Optional) Custom Base URL for self-hosted deployments
📖 Usage Examples
Chat Completion
Resource: Chat
Operation: Complete
Model: claude-4.5-sonnet
Messages: [{"role": "user", "content": "Hello!"}]
Image Generation
Resource: Image
Operation: Generate
Model: kling-v2-1
Prompt: "A beautiful sunset over mountains"
Wait for Completion: true
Video Generation
Resource: Video
Operation: Generate
Model: kling-video-o1
Prompt: "A cat playing with a ball"
Aspect Ratio: 16:9
Image → Video Workflow
- Image Node: Generate an image
- Video Node:
- Set
First Frame Binary Propertytodata - The image from the previous node will be used as the first frame
- Set
🎯 Supported Models
Chat Models
- Qwen: qwen3-235b-a22b, qwen3-max, qwen3-32b, qwen-turbo
- Claude: claude-4.5-sonnet, claude-4.5-opus, claude-4.0-sonnet, claude-3.7-sonnet
- Gemini: gemini-3.0-pro-preview, gemini-2.5-flash, gemini-2.5-pro
- DeepSeek: deepseek-r1, deepseek-v3, deepseek-v3.1
- GPT: openai/gpt-5, openai/gpt-5.2
- Others: doubao-seed-1.6, glm-4.5, kimi-k2, minimax-m2
Image Models
- Kling: kling-v2-1, kling-v2, kling-v1-5
- Gemini: gemini-3.0-pro-image-preview, gemini-2.5-flash-image
- Others: doubao-1.5-vision-pro, qwen2.5-vl-72b-instruct
Video Models
- Kling: kling-video-o1, kling-v2-1, kling-v2-5-turbo
- Veo: veo-3.1-generate-preview, veo-3.0-generate-preview, veo-2.0-generate-001
- Others: sora-2, minimax-m2, mimo-v2-flash
💾 Persistent Memory (Multi-turn Conversations)
For conversation memory that persists across workflow executions, use n8n's built-in Memory nodes:
┌─────────────────────┐ ┌──────────────────┐
│ Redis/Postgres │────▶│ Qiniu AI Agent │
│ Chat Memory Node │ │ (threadId link) │
└─────────────────────┘ └──────────────────┘
Setup:
- Add Redis Chat Memory or Postgres Chat Memory node before the Agent
- Configure the Memory node with your database credentials
- In the Qiniu AI Agent node, set the
Thread IDto match the Memory node'sSession ID - The agent will automatically resume from previous conversation context
Note: The built-in
Memorycheckpointer works within a single execution. For cross-execution persistence, use n8n's native Memory nodes.
🔗 Resources
- 📚 Workflow Templates - Ready-to-use workflow examples
- Qiniu AI SDK Documentation
- n8n Community Nodes Guide
- Qiniu Cloud AI Portal
📄 License
MIT License - see LICENSE for details.
🇨🇳 中文
n8n 社区节点 - 集成 七牛云 AI SDK 的全模态 AI 能力到 n8n 工作流中。
✨ 功能特性
| 资源 | 操作 | 描述 |
|---|---|---|
| Chat(聊天) | 文本生成 | 多模型聊天补全(通义千问、Claude、Gemini、GPT、DeepSeek 等) |
| Image(图像) | 生成、编辑 | AI 图像生成和编辑 |
| Video(视频) | 生成、混剪、查询状态 | 视频生成(可灵、Veo、Sora) |
| Audio(音频) | 文本转语音、语音转文本 | TTS 和 ASR 能力 |
| Agent(智能体) | 执行 | 支持内置工具(搜索、OCR、图像/视频生成)和 ReAct 循环的 AI 智能体 |
| Tools(工具) | 网络搜索、OCR、图片审核、视频审核、视频帧提取 | 内容安全与视频处理工具 |
🆕 v1.0.0 新功能
- 云原生状态持久化 (KodoCheckpointer):将 Agent 对话状态存储在七牛 Kodo 对象存储中,实现跨工作流执行的生产级持久化。
- 内容安全工具:
- 图片审核 (Image Censor):同步内容安全审核(涉黄、暴力、政治敏感检测)
- 视频审核 (Video Censor):异步视频内容审核,支持任务轮询
- 视频帧提取 (VFrame):从视频中提取指定时间戳的帧
- SDK v0.27.3:升级至最新 SDK,增强 Agent 能力。
📦 安装
社区节点安装(推荐)
- 进入 设置 > 社区节点
- 点击 安装
- 输入
n8n-nodes-qiniu-ai并点击 安装
手动安装
# 在 n8n 自定义节点目录
npm install n8n-nodes-qiniu-ai
🔧 配置
在 n8n 中创建凭证:
- 进入 凭证 > 新建
- 搜索 Qiniu AI API
- 输入 API Key(从七牛云控制台获取)
(可选)自定义 Base URL 用于私有化部署
📖 使用示例
聊天补全
资源: Chat
操作: Complete
模型: claude-4.5-sonnet
消息: [{"role": "user", "content": "你好!"}]
图像生成
资源: Image
操作: Generate
模型: kling-v2-1
提示词: "山间美丽的日落"
等待完成: true
视频生成
资源: Video
操作: Generate
模型: kling-video-o1
提示词: "一只猫在玩球"
宽高比: 16:9
图像 → 视频工作流
- 图像节点:生成图像
- 视频节点:
- 设置
首帧图片二进制属性为data - 上一节点生成的图像将作为视频首帧
- 设置
🎯 支持的模型
聊天模型
- 通义千问: qwen3-235b-a22b, qwen3-max, qwen3-32b, qwen-turbo
- Claude: claude-4.5-sonnet, claude-4.5-opus, claude-4.0-sonnet, claude-3.7-sonnet
- Gemini: gemini-3.0-pro-preview, gemini-2.5-flash, gemini-2.5-pro
- DeepSeek: deepseek-r1, deepseek-v3, deepseek-v3.1
- GPT: openai/gpt-5, openai/gpt-5.2
- 其他: doubao-seed-1.6, glm-4.5, kimi-k2, minimax-m2
图像模型
- 可灵: kling-v2-1, kling-v2, kling-v1-5
- Gemini: gemini-3.0-pro-image-preview, gemini-2.5-flash-image
- 其他: doubao-1.5-vision-pro, qwen2.5-vl-72b-instruct
视频模型
- 可灵: kling-video-o1, kling-v2-1, kling-v2-5-turbo
- Veo: veo-3.1-generate-preview, veo-3.0-generate-preview, veo-2.0-generate-001
- 其他: sora-2, minimax-m2, mimo-v2-flash
💾 持久化对话历史(多轮对话)
如需跨工作流执行保持对话记忆,请使用 n8n 内置的 Memory 节点:
┌─────────────────────┐ ┌──────────────────┐
│ Redis/Postgres │────▶│ Qiniu AI Agent │
│ Chat Memory 节点 │ │ (threadId 关联) │
└─────────────────────┘ └──────────────────┘
配置步骤:
- 在 Agent 节点前添加 Redis Chat Memory 或 Postgres Chat Memory 节点
- 配置 Memory 节点的数据库连接
- 在 Qiniu AI Agent 节点中,将
Thread ID设置为与 Memory 节点的Session ID一致 - Agent 将自动接续之前的对话上下文
注意: 内置的
Memorycheckpointer 仅在单次执行内有效。跨执行的持久化请使用 n8n 原生 Memory 节点。
🔗 相关链接
📄 许可证
MIT License - 详见 LICENSE。
🤝 Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
📞 Support
- GitHub Issues: Create an issue
- Email: Contact the maintainer
Made with ❤️ by bowenQT