ModelsMulti-source71°

Qwen4-Max 发布,Agent Arena 净分超 Qwen3.8-Max 7.5%

Why we picked it

阿里把 Qwen4-Max 推上 Agent Arena,净分比上代高 7.5%,实验室排名升到第 5。做智能体应用的朋友可以直接在 Model Studio 或 QwenCloud 上换 API 试试,重点关注工具调用场景的稳定性提升。

阿里巴巴发布 Qwen4-Max,在 Agent Arena 基准上相比 Qwen3.8-Max 取得 +7.5% 的净提升,使阿里整体排在第 5 位实验室。

Body

该模型通过 API 提供,接入方式包括阿里云 Model Studio 和 QwenCloud 两条渠道,同时配套 Token Plan 计费方案。对开发者来说,升级可以直接替换原有 Qwen3.8-Max 的调用端点,无需改动工作流架构。

Agent Arena 是衡量智能体任务表现的基准,这次提升集中在工具调用和多步执行能力上。阿里借此进入实验室榜单第 5 位,与 OpenAI、Anthropic 等头部实验室的差距进一步缩小,智能体方向成为各家比拼的焦点。

Original · 阿里云 Alibaba Cloud

Qwen4-Max outperforms Qwen3.8-Max by +7.5% net on Agent Arena, making Alibaba #5 lab overall.

Start building today! 🚀 - API on Alibaba Cloud Model Studio: https://t.co/D70hrCnUSx - API on QwenCloud: https://t.co/DrODwYGKXV - Token Plan on Alibaba Cloud Model Studio: https://t.co/0lfRVsAz8b - Token Plan on QwenCloud: https://t.co/t9r4DLQPqe

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Qwen4-Max, Qwen3.8-Max, Alibaba, agents.

Source attribution: curated from 阿里云 Alibaba Cloud (@alibaba_cloud). Cross-source corroboration: 11 independent sources have covered this development (including lmarena.ai, 阿里通义 Qwen, TestingCatalog), confirming industry-wide signal strength.