← November 16, 2026 briefing · added in the 16:00 KST update
Alibaba fully launches Qwen4-Max after preview, claims 62% lower cost per task and filtered-out offensive cyber skills (reports, community posts)
Announced November 16, 2026
Verified (fact)
SCMP reported that Alibaba has released Qwen4-Max, which it calls its most advanced model to date. According to the company, offensive cyber capabilities were deliberately filtered out of the training data, while capabilities useful for defense were kept. According to a GeekNews summary, the model uses a 2.4-trillion-parameter backbone built on the new Qwen4 architecture, takes images and video as direct input, and handles context of up to 1 million tokens. It also includes a 'Live Delivery' feature that first returns a usable initial result during long-running tasks and then refines it in later versions. A Reddit post said the model scored 43 on the Artificial Analysis index, below expectations. The official announcement follows a preview released on the 11th; the collected items do not confirm how the model is offered or whether its weights will be released.
Needs fact-check (claim)
Alibaba claims its score on its own Work360 evaluation rose from 47.74 for Qwen3.8-Max to 52.69. It also says lower token usage and lower API prices together cut the average cost per task by 62%, from $4.43 to $1.67 (per the GeekNews summary).
Why it matters (interpretation)
A leading Chinese open-model company released its top-tier model while highlighting a safety measure that removes cyberattack capabilities at the data stage. This suggests concern about AI-driven cyberattacks is changing product design in China's industry as well. However, some reactions say its external benchmark scores fell short of expectations, so real-world performance needs independent verification.
Confidence medium · official source pending
Sources
- Press Alibaba says new flagship AI model cuts task costs by 60% while tightening cyber guardrails South China Morning Post
- Community Qwen4-Max - 네이티브 비전과 100만 토큰 컨텍스트를 갖춘 장기 작업용 AI 모델 GeekNews
- Community Highly anticipated Qwen4 Max scores a disappointing 43 on Artificial Analysis. Reddit
Story thread
- 2026-10-29 "Qwen4 models support 1M-token context" spotted in qwen-code repo PR; Qwen4-max date appears to be a placeholder (community report, unconfirmed)
- 2026-10-29 Qwen4 follow-up: "Alibaba Cloud CTO said at Apsara Conference it will launch next week" (community report, unconfirmed)
- 2026-11-03 Qwen4 follow-up: qwen-code repo PRs list model IDs including qwen4-max, qwen4-coder-plus and qwen4-omni-120b-think (community report, unconfirmed)
- 2026-11-06 Qwen 4 follow-up: official Qwen X post previews Qwen 4 sizes, says scale will grow in Qwen 4.5 and 5 (community reports, unconfirmed)
- 2026-11-11 Qwen4 follow-up: 'Qwen4-Max Preview' released, up to 90% off on Alibaba Token Plan for a limited time (community report)
- 2026-11-12 Qwen4-Max Preview follow-up: Alibaba Token Plan only until Nov 25, 1M-token context; benchmarks and weights undisclosed (report)
- 2026-11-16 Alibaba fully launches Qwen4-Max after preview, claims 62% lower cost per task and filtered-out offensive cyber skills (reports, community posts)
- 2026-11-16 Qwen4-Max follow-up: 2.4T total parameters, 25% fewer active than predecessor; external evals show ~67% lower cost per task (Korean media)
More about Alibaba
-
Alibaba's September-quarter revenue rises 10% to 271.5 billion yuan as external AI cloud and compute recurring revenue jumps 67% (SCMP)
According to SCMP, Alibaba's total revenue for the quarter ended September 30 rose 10% year on year to 271.5 billion yuan (about $40.5...
-
Alibaba, Tencent, Moonshot AI and other Chinese firms cut models' cyberattack abilities ahead of export controls (via Bloomberg)
According to AI Times, citing Bloomberg, major Chinese AI companies including Alibaba, Tencent, and Moonshot AI have reduced...
-
Alibaba's workplace AI 'QwenWork' adds third-party models including Kimi, GLM and DeepSeek, with access control by department, group or user (Pandaily)
According to Pandaily, Alibaba has expanded the models available in QwenWork, its enterprise work tool. Besides its own Qwen models,...
-
Chinese models lead Artificial Analysis video-with-audio ranking: Alibaba's Wan 3.0 first, MiniMax H3 third, Seedance 2.5 fourth (report)
According to AI Times, Alibaba's 'Wan 3.0' ranked first with 1,175 points in Artificial Analysis's text-to-video-with-audio evaluation...
-
Qwen4-Max follow-up: 2.4T total parameters, 25% fewer active than predecessor; external evals show ~67% lower cost per task (Korean media)
According to AI Times, Alibaba's Qwen team published Qwen4-Max's architecture, benchmarks, and API pricing on its official blog on the...