News for hackers. Stuff that matters.
Re:Qwen4-Max by pella on Sunday November 15, 2026 @09:56PM (#50513480)
> Ah yes, an open-weights

This is not an open-weights modell. You have to wait for the smaller models

"Smaller models in the family are coming to open source, built on the same lineage and trained with the same generation of techniques, tuned for what small models do best."

Re:Qwen4-Max by hadlock on Sunday November 15, 2026 @10:27PM (#50513659)
Pretty excited about Qwen 4 Flash. We're running qwen 3.5 35b-a3b for low effort, rapid CRM enrichment, but Q4F will be great for more complex tasks.
Re:Qwen4-Max by electroglyph on Sunday November 15, 2026 @10:39PM (#50513742)
you can use flash next right now and it's already a massive improvement over that old 35b
Re:Qwen4-Max by bbor on Monday November 16, 2026 @09:51AM (#50518201)
...really? Isn't that massive change, if true? Wild! They're burying the lede there to say the least, if that is actually a long term stance and not just the typical "weights to come" release cycle.