跳到正文
11月19日周四
  1. 钛媒体62

    腾讯Q3财报解读:AI算力投入如何压住利润与现金流

    作者结合腾讯第三季度财报与马化腾内部讲话分析,腾讯Q3资本性支出升至651亿元,同比上涨225%,折旧摊销137亿元同比扩大56%,经营利润同比下降7%,自由现金流为-691亿元。文章认为钱主要花在GPU等算力储备上,用于混元模型升级和WorkBuddy推理,并对比谷歌搜索广告增速放缓、苹果轻投入的路径,梳理腾讯算力出租、MaaS和WorkBuddy订阅三条变现渠道。

  2. Qwen60

    Qwen 官方宣布 Qwen4-Max 已在 Nous Portal 上线,邀请用户试用并分享作品。据 Nous Research 的引用信息,这款阿里新旗舰模型与 @inferact、@Alibaba_Qwen 合作在 Nous Portal 免费开放 5 天,主打长周期任务,原生支持视觉并具备 1M 上下文。

    引用Nous Research@NousResearch

    Qwen4-Max, the new flagship model from Alibaba, is free in Nous Portal for 5 days in partnership with @inferact and @Alibaba_Qwen. It’s built for long-horizon work, with native vision and 1M context. Try it today at http://portal.nousresearch.com

  3. IT之家75

    豆包实时音视频通话模型 SeedRealtime 1.0 API 全量上线

    火山引擎正式上线豆包实时音视频通话模型(SeedRealtime 1.0)API 服务,将声音、画面、时序与表达统一到同一个端到端模型,实现边看、边听、边说全模态实时交互。

    推荐理由:原文给出端到端全模态实时交互的能力描述与多个场景演示,可据此判断实时语音应用的形态变化。

11月18日周三
  1. SenseTime60

    商汤为开源模型 SenseNova-Loop-9B 补充了更多运行方式,包括约 4.9 GB 的 4-bit AWQ 版本及配套 MTP 草稿模型,以及 Q4_K_M、Q5_K_M、Q8_0 和 BF16 四种 GGUF 格式,可在其 llama.cpp fork 中使用。两个版本已分别上线 Hugging Face 的 SenseNova-Loop-9B-AWQ 与 SenseNova-Loop-9B-GGUF 仓库。此前发布的该模型为 9B 参数,支持最高 512K 上下文和文本、图像输入。

    引用SenseTime@SenseTime_AI

    #SenseNova-Loop-9B is officially here — now #OpenSource! Just 9B parameters, competitive with Qwen3.8-27B across multiple benchmarks. 🔹 Big gains over Preview: NL2Repo 19.91 → 41.36; AutomationBench 14.33 → 36.67; GAIA 37.86 → 69.90 🔹 Loop + mHC: 32 unique layers, 56 layer executions 🔹 Day-0 4-bit AWQ weights (~4.9GB), with VRAM needs as low as ~9 GiB at 64K context 🔹 Up to 512K context, with text and image inputs See what a 9B model can build: animated posters, interactive websites and 3D scenes. 💻 GitHub: https://github.com/OpenSenseNova/SenseNova-Loop 🤗 HF: https://huggingface.co/collections/sensenova/sensenova-loop 👾 Discord: https://discord.com/invite/BuTXPHmQub

  2. Turing Post (Ksenia Se)68

    斯坦福一项论文让 OpenAI、Anthropic、Google 和 Qwen 的 8 个模型回答 620 个真实临床问题,再匿名互评,发现每个评委都偏好自己的答案,平均排名高出 1.22 位,GPT-5.6 Sol 接近 3 位。作者指出,医疗 AI 基准常用 LLM 给 LLM 打分,评委偏好自身输出会让基准偏向该模型;即使去掉评分细则,这种偏好依然存在,公开模型名称也只让偏差缩小 0.094 位。作者建议使用跨模型家族的评委小组,并按评委分别报告结果。

    引用Rohan Paul@rohanpaul_ai

    New Stanford paper finds that all 8 LLMs tested rank their own medical answers higher than other models rank them, even when answers are anonymous. Medical AI benchmarks often use LLMs to grade other LLMs' answers. If a judge favors its own outputs, the benchmark tilts toward that model. They had 8 models from OpenAI, Anthropic, Google and Qwen answer 620 real clinician questions, then rank all 8 anonymous answers. They compared where each judge placed its own answer with where the other 7 judges placed that same answer. Every judge favored its own answer, by 1.22 rank positions on average and nearly 3 for GPT-5.6 Sol. The bias held without a rubric, and revealing model names cut it by only 0.094 positions. If you use LLM judges, use a panel from several model families and report results per judge instead of 1 pooled score.

  3. 机器之心38

    UIUC等提出Ask to Distill,让大模型学会向老师提问

    UIUC等机构提出Ask-to-Distill框架,通过强化学习优化合成问题生成器,在不依赖标准答案的情况下提升知识蒸馏效率。实验显示该方法在数学推理、代码生成及智能体任务中均取得最优性能,Qwen3-4B学生模型在LiveCodeBench上成绩超越教师参考值。A2D生成的问题使回答长度减少约20%,同时累计学习信号增加约28%。

  4. 机器之心48

    港大提出 A-GRAE 打破 GRPO 隐式优势对称性,提升模型探索与准确率

    香港大学团队揭示 GRPO 存在阻碍探索与难度适应的“隐式优势对称性”,并据此提出非对称组相对优势估计算法 A-GRAE。该算法在 Qwen2.5-Math-7B 等基座模型上验证,于 AIME2025 基准将 DAPO+A-GRAE 表现提升至 53.3%,同时显著增强模型探索能力与回答多样性,代码已开源。

  5. Qwen58

    Qwen 官方宣布 Qwen4-Max 已在 Token Harbor 平台上线,定位面向真实工作场景。引用信息显示该模型主打复杂编码、多模态任务与智能体工作流,并配有 1M 上下文,Office Pass 用户可享限时 2 倍 Model Boost。

    引用Token Harbor@TokenHarborAI

    Qwen4 Max on Token Harbor Built for complex coding, multimodal tasks, and agentic workflows. ⚡ Limited-time 2× Model Boost with Office Pass. Give it a try! https://tokenharbor.ai/pricing?utm_source=twitter&utm_medium=social&utm_campaign=qwen4_max&utm_content=office_pass_boost

  6. 开源中国78

    腾讯开源 Hy4 flash 模型,单任务成本降至 Hy4 的二十分之一

    腾讯宣布开源 Hy4 flash,总参数 220B,每个 token 只激活 12B。混元团队称其为迄今最具成本效益的 Hy 模型,单任务平均成本只有 Hy4 的 1/20、DeepSeek-V4.1-Flash-1017 的 1/8。相比上一代 Hy4,该模型在结构上改了三处。

    同一件事,精选收录的是《腾讯混元开源 Hy4 flash 模型,220B 总参数、12B 激活、支持 1M 上下文》

  7. 腾讯混元模型发布82

    腾讯混元开源 Hy4 flash 模型,220B 总参数、12B 激活、支持 1M 上下文

    腾讯混元团队开源 Hy4 flash,总参数 220B(含 64B Engram 参数),每 token 仅激活 12B,原生支持文本与图像输入、输出文本,上下文窗口达 1M tokens。

    推荐理由:官方给出总参数 220B、激活 12B 的架构细节与每任务成本对比,可据此判断轻量 MoE 在生产力任务上的取舍。

  8. 雷锋网52

    千问办公电商版推出品牌知识库功能

    11月18日,千问办公电商版推出“品牌知识库”功能,商家在网页端切换至“电商经营”模式并上传网店链接、过往图片和设计规范后,生成图片、视频和文案时会统一品牌调性。该功能包含品牌档案、品牌参考集和品牌记忆三项能力,品牌记忆会在后续对话中自动调用品牌背景。此前该功能面向部分商家邀测,目前已向所有电商商家开放。

  9. Qwen75

    Qwen4-Max 已在 Runware 上线,支持文本、图像和视频输入,上下文窗口为 1M token,输出上限 128K token。该模型支持可配置的推理强度、function calling 和 JSON Schema 输出,思考与非思考模式同价,输入 1M token 以内为单一档位。可通过 /v1/chat/completions 或 /v1/responses 调用 alibaba:qwen@4-max,定价为每 1M 输入 token 1.50 美元、输出 4.50 美元、缓存输入 0.075 美元。

    引用Runware@runware

    @Alibaba_Qwen Qwen4-Max is live on Runware 🧠 Alibaba's Qwen4 flagship takes text, images and video in a 1M-token context window, with configurable reasoning effort, function calling and JSON Schema output. - 1M-token context, up to 128K output tokens - Image and video input for screenshot, document and video analysis - Same price in thinking and non-thinking modes, one input tier up to 1M tokens - Call alibaba:qwen@4-max on /v1/chat/completions or /v1/responses from $1.50 per 1M input tokens, $4.50 output and $0.075 cached input

    推荐理由:Qwen4-Max 接入 Runware 后,读者可据此了解其 1M token 上下文与多模态输入的实际调用方式和定价。

11月17日周二
  1. SenseTime76

    商汤正式开源 SenseNova-Loop-9B,仅 9B 参数,官方称在多项基准上与 Qwen3.8-27B 相当。相比 Preview 版本,NL2Repo 从 19.91 提升到 41.36,AutomationBench 从 14.33 提升到 36.67,GAIA 从 37.86 提升到 69.90。

    推荐理由:9B 参数模型在多项基准上对标更大模型,并给出 4-bit 权重和 512K 上下文的部署条件,便于评估小模型的实际可用边界。

  2. Qwen58

    Qwen4-Max 已在 LebrelAI 平台上线,用户可试用其智能体编码能力。引用信息称该模型在 DeepSWE v1.1 上排名第一,领先 Kimi、GPT 和 Opus,并以低于阿里官方定价 7% 的价格提供,每条回答附带签名回执。

    引用LEBREL AI@LebrelAI

    The top agentic coding model, live on Lebrel at 7% below Alibaba's list price. Qwen4 Max from @Alibaba_Qwen ranks first on DeepSWE v1.1, ahead of Kimi, GPT and Opus, per Qwen's results. Anonymized, attested, with a signed receipt on each answer. Try it: https://Lebrel.ai

  3. 钛媒体42

    AI4S火热,离“ChatGPT时刻”还有多远?

    阿里发布Qwen4-Max展示自主设计新型冷却液分子并调用自动化实验平台完成合成与测试的能力。Anthropic的Claude Fable 5.2和Mythos 5.2、腾讯混元Hy4正式版也相继展示了通用大模型解决理论物理、数学研究等领域复杂问题的能力。

  4. 机器之心78

    Prefrontal AI 与 ETH Zurich 提出 LSA:10M token 下比 DeepSeek 式稀疏注意力再快 22.7 倍

    伦敦初创公司 Prefrontal AI 与苏黎世联邦理工学院联合发布论文,提出隐空间稀疏注意力 LSA,在 10M token 下把 GLM-5.3-Flash 单个注意力层的 prefill 提速 22.7 倍、Qwen3.8-27B 提速 72.4 倍,精度与原模型相近。

    推荐理由:论文指出超长上下文下索引器占 prefill 时间 98%,并给出按语义聚类把索引降到 O(T log T) 的路径。

  5. 智东西60

    百度三季度营收308亿元,AI业务收入占比达53%

    百度公布2026年第三季度财报,营收308亿元同比下降1%,AI业务收入132亿元占一般性业务收入的53%,已连续三个季度占比超50%。其中AI云基础设施收入82亿元、GPU云收入同比增长191%,AI应用收入26亿元与去年同期持平,AI原生营销服务收入24亿元同比下降15%。李彦宏称昆仑芯今年年初已秘密提交上市申请,百度正持续推进上市流程。

  6. Tencent Hy42

    Tencent’s newly launched AI model is impressive! 🤯 I tested the generation speed of Hy4 Flash using a variety of prompts. It took less than 15 minutes to generate this 3D game scene. Plus, the cost was over ten times lower compared to the previous Hy4 model! You can try it for free on WorkBuddy until November 30. Go check it out! @TencentHunyuan @WorkBuddy_AI

    引用Rebecca Adson@RebeccahAdson

    Tencent’s newly launched AI model is impressive! 🤯 I tested the generation speed of Hy4 Flash using a variety of prompts. It took less than 15 minutes to generate this 3D game scene. Plus, the cost was over ten times lower compared to the previous Hy4 model! You can try it for free on WorkBuddy until November 30. Go check it out! @TencentHunyuan @WorkBuddy_AI

  7. Tencent Hy38

    🚨 Over the past couple of days, I’ve had the chance to test Tencent’s new Hy4 Flash model and compare it with the previously released Hy4 model. I created this game using the latest Hy4 Flash AI model. ⏰ Time taken: 63 minutes The game includes: 👉 Swords 👉 Rockets 👉 Medkits 👉 Shields The game mechanics are impressive! 👀 Check out the 0:37 mark. A soldier actually uses a shield to block an arrow I fired! 🤯 I then ran the same prompt through the Hy4 AI model again. ⏰ Time taken: 90 minutes It’s worth noting that the 3D world and character designs in this version look even more realistic. In short, Hy4 Flash is faster and more cost-effective for handling tasks, while the full Hy4 model is better suited for more complex requirements. Hy4 Flash is also available for free on the WorkBuddy platform until November 30. Go check it out! @TencentHunyuan @WorkBuddy_AI

    引用KATE@kate_osita_

    🚨 Over the past couple of days, I’ve had the chance to test Tencent’s new Hy4 Flash model and compare it with the previously released Hy4 model. I created this game using the latest Hy4 Flash AI model. ⏰ Time taken: 63 minutes The game includes: 👉 Swords 👉 Rockets 👉 Medkits 👉 Shields The game mechanics are impressive! 👀 Check out the 0:37 mark. A soldier actually uses a shield to block an arrow I fired! 🤯 I then ran the same prompt through the Hy4 AI model again. ⏰ Time taken: 90 minutes It’s worth noting that the 3D world and character designs in this version look even more realistic. In short, Hy4 Flash is faster and more cost-effective for handling tasks, while the full Hy4 model is better suited for more complex requirements. Hy4 Flash is also available for free on the WorkBuddy platform until November 30. Go check it out! @TencentHunyuan @WorkBuddy_AI

  8. Qwen70

    Qwen 官方宣布 Qwen4-Max 已在 You.com 上线,可通过其 Inference Router 调用。据 You.com 介绍,这是 Qwen4 系列的首个模型,面向编码、办公任务和长程智能体运行,支持 1M-token 上下文以及原生文本、图像和视频输入。

    引用You.com@youdotcom

    Qwen 4-Max is live on Inference Router, from Day 0. The first model in @Alibaba_Qwen's Qwen4 series, built for coding, office work, and long-horizon agent runs. 1M-token context, native text, image, and video input. Try it now: https://you.com/product/inference-router/models/qwen4-max

    推荐理由:Qwen4 系列首个模型上线第三方推理路由,给出了上下文与多模态输入的规格,可据此判断其适用场景。

11月16日周一
  1. Ethan Mollick78

    Ethan Mollick 表示,Fable/Mythos 首次公开已超过 8 个月,但至今没有出现同级别的开源模型,这与许多人的预期相反。他引用 Artificial Analysis 的数据:阿里发布 Qwen4 Max,在 Intelligence Index 上得分 43,比 Qwen3.8 Max 高 9 分,比 Kimi K3.1 Pro 低 3 分,每任务成本约低 11%。

    引用Artificial Analysis@ArtificialAnlys

    Alibaba has launched Qwen4 Max, gaining 9 points over Qwen3.8 Max on the Artificial Analysis Intelligence Index and landing 3 points behind Kimi K3.1 Pro at ~11% lower Cost per Task Qwen4 Max, using xhigh reasoning effort, scores 43 on the Intelligence Index, 9 points ahead of Qwen3.8 Max (xhigh) and 3 points behind Kimi K3.1 Pro (max, 46) Key takeaways: ➤ Qwen4 Max improves on every evaluation in the Index vs Qwen3.8 Max (xhigh, 34): HLE-Diamond +21 points (24% to 45%), AA-Omniscience +18 points, AutomationBench-AA +17 points, CritPt v2 +13 points, Terminal-Bench-Science +11 points, Terminal-Bench 4.0 +9 points, GDP.pdf +3 points, FrontierCode 2.0 +3 points and AA-Briefcase +82 Elo. ➤ Qwen4 Max costs $3.27 per Intelligence Index task, ~3x less than Qwen3.8 Max ($10.02) and ~11% less than Kimi K3.1 Pro ($3.68): This is driven by 25% lower per-token pricing than Qwen3.8 Max, 70% cheaper cache hits and lower token usage. Tencent Hy4 costs ~3x less per task ($1.03) than Qwen4 Max but scores 8 points lower on the Index. ➤ Qwen4 Max uses 39% fewer output tokens per task than Qwen3.8 Max: ~105k output tokens per Intelligence Index task vs ~171k for Qwen3.8 Max, ~72k for Kimi K3.1 Pro and ~78k for Tencent Hy4. Qwen4 Max uses fewer tokens, while increasing 9 points on the Intelligence Index ➤ Improvements in AA-Briefcase (knowledge work): On AA-Briefcase, our evaluation of agentic work across real business files, Qwen4 Max reaches 1686 Elo vs 1674 for Kimi K3.1 Pro and 1446 for Tencent Hy4. ➤ Qwen4 Max trails Kimi K3.1 Pro on Terminal-Bench 4.0 and Tencent Hy4 on FrontierCode 2.0: Qwen4 Max scores 48% on Terminal-Bench 4.0 vs 67% for Kimi K3.1 Pro and 37% for Tencent Hy4. On FrontierCode 2.0, @cognition’s private evaluation of real-world coding tasks, it scores 26%, level with Kimi K3.1 Pro (26%) and behind Tencent Hy4 (31%). Key model details: ➤ Context window: 1M tokens ➤ Multimodal: Text and image input with text output ➤ Pricing: $1.50/$4.50 per 1M input/output tokens on Alibaba's first-party APIs, with a $0.075 cache hit price. This is lower than Qwen3.8 Max across the board ($2.00/$6.00, with a $0.25 cache hit price) ➤ Availability: @Alibaba_Qwen 's first-party APIs

    推荐理由:作者以开源模型发布节奏为参照,指出 Fable 级开源模型迟迟未出现,可与引用中的闭源模型进展对照阅读。