AI 模型路由插件model-router里面的自动匹配模型多模态不支持的问题 || The automatic matching model multi-modality in the AI model routing plug-in model-router does not support the problem
Author: zpj838Created Sep 17, 2026Updated Sep 18, 2026
AI 模型路由插件model-router里面的自动匹配模型功能，autoRouting 只匹配最后一条 role=user 消息的纯文本部分，看图纸type:image的消息不会匹配，这个有什么办法让页能按照type:image消息去匹配吗？我的场景是如果检测到附件中有图片自动路由到支持图片解析的指定模型下 。
The automatic matching model function in the AI model routing plug-in model-router, autoRouting only matches the plain text part of the last role=user message, and the drawing type:image message will not match. Is there any way to make the page match according to the type:image message? My scenario is that if an image is detected in the attachment, it will be automatically routed to a specified model that supports image parsing.
The automatic matching model function in the AI model routing plug-in model-router, autoRouting only matches the plain text part of the last role=user message, and the drawing type:image message will not match. Is there any way to make the page match according to the type:image message? My scenario is that if an image is detected in the attachment, it will be automatically routed to a specified model that supports image parsing.
Source: higress-group/higress