6/18/2026 at 3:58:09 PM
by Brajeshwar
6/18/2026 at 9:26:06 PM
It’s interesting how little press Minimax M3 gets, given it outperforms Deepseek V4 Pro, previously the SOTA for open models. Meanwhile GLM has been in the news daily.by ricardobeat
6/19/2026 at 3:43:11 AM
It is strange, huh? But the hype cycles around these models often ignore good contenders. Xiaomi's MiMo-V2.5 Pro was doing really well and didn't get much hype either.by Reubend
6/18/2026 at 4:25:50 PM
I wonder if multiple attempts at the opossum would produce better results.If we didn’t have the previous example I would interpret this as pretty solid evidence that labs were training on the Pelican “benchmark”.
I just can’t imagine a model dropping so significantly from one version to the next on such a silly task.
by besterman23
6/18/2026 at 8:26:09 PM
Related:GLM-5.2 is the new leading open weights model on Artificial Analysis
by ChrisArchitect