Y
Hacker News
new
|
ask
|
show
|
jobs
by
darknoon
305 days ago
It's ok, somewhere between a qwen 2.5 VL and the frontier models (o3 / opus 4) on visual reasoning