I selected a large model node. For its context, I chose the output of the node before it, but this output is of type Array\[Object\]. An error was reported during runtime. Then I wanted to select the output of the previous node in the system prompt, but found that the previous node’s information did not appear in the reference panel at all, meaning it couldn’t be referenced. It could only be referenced via “context”, but it still reported an error when actually run. I don’t know what the underlying reason is!
After the image upload is complete, it’s just a blank image. What the heck? I’ve tried uploading it several times, both from the clipboard and local files.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Dify 1.11.2配置gemini和gpt的模型后工作流报错 | 6 | 74 | January 16, 2026 | |
| LLM和Agent节点的上下文变量怎么设置,为什么是空的 | 1 | 117 | January 12, 2026 | |
| Gemini 模型插件报错:PluginInvokeError | 7 | 110 | January 21, 2026 | |
| Dify对话运行时报错,显示 运行失败 | 1 | 25 | January 26, 2026 | |
| 大模型输出了mermaid代码块,但是在dify平台无法正常渲染成图形 | 1 | 42 | February 4, 2026 | |
| 为什么我的直接回复的节点内容和在预览界面的输出内容不一致 | 2 | 61 | January 8, 2026 | |
| 为什么无法调用大模型,不管选哪个都报错 | 0 | 97 | December 10, 2025 | |
| Array[File],File 的合并问题,有大神吗? | 2 | 155 | December 16, 2025 | |
| Dify chatflow 最后的直接回复节点引用了开始节点的files | 1 | 39 | January 13, 2026 | |
| Llm节点如何支持多模态输入? | 6 | 224 | January 16, 2026 |
