MaxxBench – "Nice DeepSWE score. But is it true?"

Re:MaxxBench – "Nice DeepSWE score. But is it true?" by s1artibartfast on Sunday November 08, 2026 @02:27AM (#50411145)

Users always have the choice to ignore benchmarks or run their own

[ Parent | on HN ]

Re:MaxxBench – "Nice DeepSWE score. But is it true?" by danuker on Sunday November 08, 2026 @02:40AM (#50411192)

Exactly. Even a brief one (a single task done on two models) showed me GLM 5.3 Flash was superior to DeepSeek 4.1 Flash.

DS used nearly twice the budget to accomplish the same thing.

Ideally, these tasks should be something you actually need day to day.

[ Parent | on HN ]

Re:MaxxBench – "Nice DeepSWE score. But is it true?" by lisplist on Sunday November 08, 2026 @11:53PM (#50421211)

I'd rank it DeepSeek 4.1 Flash 1017 > GLM 5.3 Flash > DeepSeek 4.1 Flash (Original) - 1017 is a huge improvement.

[ Parent | on HN ]

Re:MaxxBench – "Nice DeepSWE score. But is it true?" by danuker on Monday November 09, 2026 @04:19PM (#50432326)

Oh! I thought I had tested 1017, but I forgot about it. Thank you!

[ Parent | on HN ]

Re:MaxxBench – "Nice DeepSWE score. But is it true?" by adastra22 on Sunday November 08, 2026 @04:55AM (#50411816)

That’s missing the point. Frontier models get shittier in certain ways with each release, due to benchmaxxing, and there is nothing we can do about that.

[ Parent | on HN ]