ModelBrief · Qwen4-Max-Preview · Topic Page
Last updated Nov 13, 2026

Cross-Vendor: Qwen4-Max vs Claude, Gemini and GPT — Why It Cannot Be Ranked Yet

Published Nov 13, 2026 · Last updated Nov 13, 2026

The question every reader asks: how does Qwen4-Max stack up against the Claude, Gemini and GPT flagships? Our answer may disappoint, but it is honest: it cannot be ranked yet, because it does not have a single verifiable score. Here is everything that is known, and why it is not enough.

The shared table: one blank cell

ModelAPI price $/1M in/outAA indexContext
Qwen4-Max-PreviewToken Plan subscription onlyNot listed1M
Claude Fable 5.28 / 4065 (max)1M
Gemini 4 Carbon2 / 10 (until Nov 30)60 (high)1M
GPT-6 Astra10 / 5057 (max)1.05M
GPT-6.1 Sol2 / 1055 (max)1.05M
Qwen3.8 Max2 / 645 (Sep 27)1M

Source: Quantslant, Nov 11, 2026; AA index = Artificial Analysis Intelligence Index.

What head-to-head anecdotes can and cannot say

With scores absent, only same-prompt anecdotes remain: one tester rated its space shooter above Fable 5.2's and Carbon's; in a T-Rex game test it finished faster while Fable 5.2 was still judged best overall; a Hacker News user found it slightly slower than GPT-6.1 Terra. These point in contradictory directions — which is precisely the proof that anecdotes cannot substitute for benchmarks. Use them to decide whether it deserves your own test, never to declare a ranking.

About the circulating "53" chart

A chart claiming an AA index of 53 for Qwen4-Max circulates on social media. As documented in this site's rumor-check section, the score appears nowhere in AA's own leaderboard or posts, and the chart contradicts its own legend: it is unverified and should not be cited. When a real score exists, this page will be updated.

Our take: on price structure and context, Qwen4-Max sits at the same table as every flagship here; on capability ranking, it is temporarily absent. Until Artificial Analysis or LMArena publishes a formal score, any claim of the form "Qwen4-Max ranks Nth" is fabrication. Our advice stands: test it on your own work now, read the leaderboard later.

Reference source: this article is an original rewrite produced by this site; facts and figures were checked against Quantslant 汇总(Artificial Analysis 指数)(2026-11-11)

← Back to the Qwen4-Max topic page