my manus never call vision llm

[castleKing1997]

I have configured qwen-vl-plus as the vision llm and clearly specified in the task description that Manus should use vision llm in the task. However, this model has not been called as expected.

[AbdullaevG]

I have the same problem, I specified both LLM and VLM in the config file, but requests to VLM do not go at all

[carrot-codeing]

same question, i have llm and llm.vision config in confif.toml, but the vlm never called. @fred913

[BahamutRU]

The same. Up!

Bot: I cannot directly view or evaluate images. Would you like me to list the image filenames in the folder instead, or should I terminate the task?