#KAT-Coder-Pro V2
11 messages · Page 1 of 1 (latest)
Didn't see anyone talking abt this, but it's on OpenRouter now
pretty okay results on coding benchmarks, but what makes this interesting is that it isn't a reasoning model, so you'll be paying less overall for tokens
currently priced the same as Minimax M2.7
it should be roughly half the price of M2.7 if I can count and the recent activity on openrouter is accurate to most conversations
No reasoning means faster speed too? We'll see
Idk why qwen 3 coder next didn't take off, maybe this will
theoretically no reasoning means responses are almost instant
I just fear that for complex tasks it may not fare so well
vibes why how does it compare with other chines models?
Nobody is talking about it because their last model was benchmaxxed waste of gpu shit