← Tous les modèles

Qwen4-Max

Pas la peine
Chat· $0.075/M input (cache hit)· Dernière mise à jour nov. 2026Site officiel

Alibaba's 2.4T-parameter MoE flagship (71B active), native vision, 1M-token context — 6th on Code Arena.

Ce que dit la communauté

👍 46%·14 reviews
  • mixed

    “Tried it in Qwen Code; with xhigh thinking the speed is okay, feels slightly slower than Terra. ... In practice, this discount among Chinese models (those usable for production) should currently be among the most durable tier (and it only has a monthly quota).”

    hn ↗

  • mixed

    “It's not bad — backend ability is OK, and since it's a new foundation model there's still a lot of RL headroom. Domestic models aren't far behind SOTA on backend now, which shows the coding ability is there; the frontend gap comes more from clearly weaker tool calling and multimodality.”

    bilibili ↗

  • positive

    “Judging from just these few tests, it's the best among the domestic models.”

    bilibili ↗

FAQ

Qwen4-Max vaut-il le coup ?

Plutôt non — Qwen4-Max n'a qu'un score worth-it de 47%, basé sur 116 votes de la communauté.

Dans quoi Qwen4-Max excelle-t-il ?

Qwen4-Max est un modèle de Chat. 2.4T total parameters with only 71B active per token — lower inference cost at the same scale.

Combien coûte Qwen4-Max ?

$0.075/M input (cache hit)

avis (0)

Pas encore d'avis sur ce modèle.

Worth it in this category

More →
47%Ça vaut le coup