Re:Qwen4-Max
by m00dy on Sunday November 15, 2026 @10:15PM
(#50513602)
I bet most of us waiting for 27b model and want to see how good it will be.
[ Parent
| on HN ]
Re:Qwen4-Max
by McGlockenshire on Sunday November 15, 2026 @10:18PM
(#50513621)
My 3060 yearns for the next generation MoE!
Re:Qwen4-Max
by kylehotchkiss on Sunday November 15, 2026 @10:28PM
(#50513665)
M5U locked and loaded.
I wish there were more ~60b dense options like ripple too!
I really want a M5U but man that price is hard to justify for a base model when I paid less than half that for my 96GB macbook.
Do you run Flash Next or 27b on yours?
Re:Qwen4-Max
by kylehotchkiss on Monday November 16, 2026 @12:07PM
(#50519912)
Oh, I leased it :D a trade-in helped reduce the cost a little too. I recommend that route. It's doing well over $100/month in inference value for me and the CPU is still totally free while it's working on that and is a great computer for everyday life too.
I'm actually running Gemma4 for the most part - for my specific needs it was the best out of the several I evaled
Re:Qwen4-Max
by neutrinobro on Sunday November 15, 2026 @10:58PM
(#50513846)
Spooling up the V100s right now...
I’m a QFN dude myself, but yes can’t wait to get it loaded on a spark.
Forget about 27b when are we getting new 4b, 9b and moe qwens, i'm dying out here
what about Bolt? so far I enjoy it alot on my NAS
Re:Qwen4-Max
by kylehotchkiss on Monday November 16, 2026 @12:39PM
(#50520327)
What are you using the smaller models for?