Models73°
阿里更新旗舰模型Qwen3.8-Max
阿里更新旗舰模型Qwen3.8-Max，前端编程能力跃居全球第一
Why we picked it
阿里Qwen3.8-Max编程能力登顶全球，代码生成超越GPT-4o，开发者必看。
阿里发布Qwen3.8-Max模型，在HumanEval编程基准上得分为84.5分，超越GPT-4o成为全球第一。该模型在代码生成、调试和优化方面表现优异，支持多种编程语言。Qwen3.8-Max还增强了多模态处理能力，能够理解和生成图文内容。
Keep reading
Recent stories with the same tag or sectionAlibaba Token Hub Shows Agentic AI Workflows at Apsara Conference 2026
The talk signals Alibaba Cloud's push to position its Qwen family as agent-native infrastructure rather than chat models, a framing that affects how enterprises in Singapore and Southeast Asia choose MaaS providers. Teams evaluating Alibaba Cloud against regional cloud rivals can watch the session to compare its agent stack and realtime audio capabilities. Builders planning multi-step agent workflows on Qwen models will find the execution-oriented framing directly relevant.
Chinese content英伟达投资的 Reflection AI 发布首款开放权重模型 Beam
英伟达投的 Reflection AI 憋了个大招，Beam 开放权重，直接叫板 DeepSeek 和 GLM-5.2，跑代码和智能体任务的朋友可以关注下。
Chinese contentQwen3.8-Max 修复 PVE 服务器故障
Qwen3.8-Max 实际修复服务器案例，对比 Grok-4.7 和 DeepSeek-V4.1-Flash 表现更优。
Chinese contentQwen3.8-Max 完成 33 轮自动迭代，Artificial Analysis 分数从 40 升到 45
阿里 Qwen3.8-Max 被曝跑了 33 轮自动训练迭代，AA 分数从 40 涨到 45，还有 2.4T 参数的说法。
Qwen3.8-Max showcased at Apsara Conference 2026
An independent vLLM maintainer publicly backing Qwen3.8-Max signals that Alibaba's flagship model is being taken seriously by the open-source inference community that most developers rely on for serving. Teams evaluating agentic stacks can treat this as a data point when comparing Qwen against DeepSeek or open-weight Llama alternatives. Follow-ups from the conference may include more Qwen updates.
Chinese contentAlibaba 透露 Qwen4 已进入训练，路线图指向 5–10T 参数的 Qwen4.5 和 Qwen5
Alibaba 在 Apsara 2026 透露 Qwen4 正在训练，后续 Qwen4.5、Qwen5 要做到 5–10T 参数，用 Qwen 的可以关注下节奏。