trying to use glm-4.6v (from zai provider) for image and video understanding
quick question about media understanding config
i'm trying to use glm-4.6v (from zai provider) for image and video understanding. i already have zai configured and working for text with the ZAI_API_KEY
is there a way to set it up as a media.models entry directly using the zai provider? or do i need to use the cli fallback approach with the zai command?
or does it have to be:
thanks
i'm trying to use glm-4.6v (from zai provider) for image and video understanding. i already have zai configured and working for text with the ZAI_API_KEY
is there a way to set it up as a media.models entry directly using the zai provider? or do i need to use the cli fallback approach with the zai command?
{
"provider": "zai",
"model": "glm-4.6v",
"capabilities": ["image", "video"]
}
or does it have to be:
{also wondering if anyone has gotten glm vision models working for media understanding and what the exact config looks like
"type": "cli",
"command": "zai",
"args": ["-m", "glm-4.6v", "{{MediaPath}}"],
"capabilities": ["image", "video"]
}
thanks
Solution
Great question! For the ZAI provider with glm-4.6v, here's what you need to know:
Direct Provider Approach (if supported):
CLI Fallback (if provider doesn't support media directly):
Direct Provider Approach (if supported):
CLI Fallback (if provider doesn't support media directly):
