Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving

Qwen dropped a preview of its new flagship AI and the internet did what it does best: started a comment cage match. The company says Qwen3.6‑Max‑Preview has sharper world knowledge, follows instructions better, and writes code more reliably than its last model — and you can try it on Qwen Studio. But skeptics swarmed the benchmarks. One user blasted the comparisons for using older “Opus 4.5” instead of newer versions, while another asked why OpenAI wasn’t even in the charts and why a Chinese model, “GLM 5.1,” was suddenly the star. Translation: cue the “benchmark bingo” memes and side‑eye at cherry‑picking. Then came the identity crisis: some folks know Qwen’s free, downloadable models, yet had no clue about the paid cloud lineup — a brand split that left people asking, “So… what am I supposed to use?” Meanwhile, the value warriors showed up. One bragged he ships all his code with a cheaper rival for $10 a month, dunking on the whole state‑of‑the‑art race. And the big mood? Fear that shiny previews mean fewer open, freely available AIs. The open‑weights crowd pleaded: don’t lock it all up. In short: new model, same internet — hype vs. receipts vs. wallet