Alibaba's Qwen3.8-Max barely edges DeepSeek-V4-Flash — at 20x the price

Alibaba has officially released Qwen3.8-Max, a 2.4-trillion-parameter multimodal flagship model aimed at coding agents, professional office work, and long-running tasks. The API is now live.

On four benchmarks — TerminalBench 2.1, DeepSWE 1.1, NL2Repo, and Toolathlon Verified — it beats DeepSeek-V4-Flash by just 1.7 to 3.9 points. The two companies run different test frameworks, so this isn't a strict head-to-head, but the margin is undeniably thin. The price gap, however, is not: Qwen3.8-Max costs nearly 20 times more.