Alibaba released Qwen 3.8-Maxthis week and marketed the preview as second only to Claude Fable 5 (theirlaunch-day tablewas more equivocal: the model leads on one of 12 coding-agent rows). But an independent harness came close to the opposite conclusion: abenchmark run, apparently using the Preview version, put Qwen 3.8-Max's best effort setting mid-pack, and its default setting last.Both results a
Alibaba released Qwen 3.8-Maxthis week and marketed the preview as second only to Claude Fable 5 (theirlaunch-day tablewas more equivocal: the model leads on one of 12 coding-agent rows). But an independent harness came close to the opposite conclusion: abenchmark run, apparently using the Preview version, put Qwen 3.8-Max's best effort setting mid-pack, and its default setting last.Both results a