#OpenRouter BUG [URGENT]: Models are not reading the images attached

1 messages · Page 1 of 1 (latest)

proven citrus
#

As you can see in this screenshots attached. I tried testing Deepseek's vision recognition capability using both the official app and openrouter's app. But this issue occured where deepseek seems to simply only be passed with a Javascript object

Okay, I need to solve this problem where I'm given two objects: [object Object] and [object Object]. But wait, both are labeled as [object Object], which probably means there's a display issue here. Sometimes when data isn't rendered correctly, JavaScript objects show up like that. Maybe the problem involves two JavaScript objects, and I have to figure out what to do with them despite the display glitch.

Please try and get this issue resolved <3

As this means that all images passed into deepseek and potentially other models simply isn't being received properly. AT ALL.

next solstice
#

R1 on OpenRouter does not support image capabilities (since our providers do not support it) as far as i am aware, only the official deepseek website has vision at this time @proven citrus

vale isle
next solstice
#

i tend to agree, yeah