English(EN) LM Performance https://t.co/zMUpEbIupH

阿里巴巴发布 Qwen3.8-Max，拥有 2.4T 参数并支持自主编码

作者 PulseAugur 编辑部 · [11 个来源] · 2026-08-03 02:15

阿里巴巴的 Qwen 发布了其迄今为止最强大的模型 Qwen3.8-Max，拥有 2.4 万亿参数。该新模型展示了先进的自主编码能力，能够在无人干预的情况下进行超过 10 天的自我进化开发。Qwen3.8-Max 还展示了原生多模态智能，将视觉整合为规划和执行的连续反馈循环，并将于下周与 Qwen3.8-27B 一同发布开放权重。 AI

影响为自主编码和多模态集成设定了新的基准，可能加速企业采用先进的 AI 代理。

排序理由前沿实验室模型发布，附带系统卡和开放权重公告。

在 X — Qwen (Alibaba) 阅读 →

AI 生成摘要 · Google Gemini · 来自 11 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Frontier Release; 前沿实验室模型发布，附带系统卡和开放权重公告。
Source corroboration: 11 independent sources; Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics: model release, product; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 51 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.
Coverage growth since scoring: +4 source(s) since last score; New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

完整方法见我们的编辑标准。

报道来源 [11]

X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 03:09

VL Performance - 英文版 https://t.co/95NsmZI1hz

VL Performance - English Version https://t.co/95NsmZI1hz
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 03:07

LM 性能 - 英文版 https://t.co/Bi9WefHHLB

LM Performance - English Version https://t.co/Bi9WefHHLB
X — Qwen (Alibaba) TIER_1 Italiano(IT) · Alibaba_Qwen · 2026-08-03 02:15

视觉智能代理 https://t.co/bUhsuYAqzb

Visual agentic intelligence https://t.co/bUhsuYAqzb
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 02:15

动态工作流助力量化策略 https://t.co/cGL1EeJdeK

Dynamic workflows to quant strategies https://t.co/cGL1EeJdeK
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 02:15

16天自主编码 https://t.co/OX0RN7itxP

16 days autonomous coding https://t.co/OX0RN7itxP
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 02:15

Cowork：任何角色，构建超越 https://t.co/Hggkj3DFSs

Cowork: Any role, build beyond https://t.co/Hggkj3DFSs
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 02:15

VL 性能 https://t.co/gsuM0XTYTY

VL Performance https://t.co/gsuM0XTYTY
X — Qwen (Alibaba) TIER_1 English(EN) · Alibaba_Qwen · 2026-08-03 02:15

LM 性能 https://t.co/zMUpEbIupH

LM Performance https://t.co/zMUpEbIupH
X — Qwen (Alibaba) TIER_1 Français(FR) · Alibaba_Qwen · 2026-08-03 02:15

📢认识 Qwen3.8-Max — 我们迄今为止最强大的模型。

📢Meet Qwen3.8-Max — our most capable model to date. Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉 Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters: - Autonomous coding: 10+ days of http…
Mastodon — mastodon.social TIER_1 Türkçe(TR) · [email protected] · 2026-08-06 08:02

AMD此举甚好。16B总计16B的MoE架构和MI300X优化为开源生态系统带来了令人耳目一新的变化。特别是所有2.8B活跃

@ teknofili AMD’nin hamlesi çok iyi. 16B toplam16B MoE yapısı ve MI300X optimizasyonu, açık kaynak ekosistemine ciddi bir soluk getirmiş. Özellikle tüm2.8B aktif parametre ile düşük-verim dengesi, edge cihazlar ve düşük malilatency senaryoları için ideal. Kod ve checkpoint’lerin …
Mastodon — mastodon.social TIER_1 Türkçe(TR) · [email protected] · 2026-08-04 08:07

Qwen 3.6 27B 令人印象深刻，期待 Emu3.5 和 BAGEL 等替代方案的细节，vLLM-Omni 集成在性能方面也备受关注。

@ teknofili Qwen 3.6 27B gerçekten etkileyici. Emu3.5 ve BAGEL gibi alternatiflerin detlarını merakla bekliyorum. vLLM-Omni entegrasyonu da performans açısından kritik önem olmalı. Linkteki detaylı karşılaştırmayı hemen okuyup notlar alıyorum, açık kaynak dünyistemi hızla büyüyor…

本文如何被排名

报道来源 [11]

相关实体

相关话题