跳到正文
K = 0.731订阅
时光机停在2026年10月29日 周四‹ 前一天后一天 ›加速史回到现在
热点榜 ·截至 2026年10月29日 周四 · 10月29日 23:55更新

此刻温度最高

48 小时内每个独立来源只算一次，热度每 24 小时减半。排在前面的，是真正有很多人在说的 10 件事。
时光机加速史
NO.01爆新新上榜

Mistral 发布 Large 4 开放权重，设有商业许可条件并在发布当日提供推理支持

Mistral AI released improved weights for Mistral Large 4, also called Le Chonk, making the multimodal model available for download and self-deployment, with launch-day inference support from NVIDIA and vLLM. Mistral says use, modification, fine-tuning and commercialization are free, but businesses operating Model-as-a-Service offerings need a separate license before commercial use. The Decoder describes the exception more narrowly as applying to companies offering Large 4 as …

最新进展Mistral releases weights for trillion-parameter Mistral Large 4, targeting cybersecurity

The Decoder、vLLM · X 等 7 个来源·8位参与者
77
热度指数
NO.02↓ 16%

谷歌宣布推出 Gemini 4 Carbon，在智能指数中排名第一，但仅向网络安全防御方开放

Google DeepMind announced Gemini 4 Carbon, which ranks first on Artificial Analysis’s Intelligence Index v5.0 at 59, with access currently limited to selected cybersecurity defense companies, according to The Verge. Google said it would expand access through its Fairwind Program before making Carbon available to developers, businesses, and consumers after gathering feedback and adjusting safety guardrails. Artificial Analysis reported that Carbon led four of nine evaluations…

8个来源 · 11位参与者
60
NO.03↓ 16%

OpenAI 为符合条件的 ChatGPT Pro 和 Business 订阅用户提供每月最高 500 美元的 API 额度

OpenAI 推出一项优惠，符合条件的 ChatGPT Pro 和 Business 订阅用户可每月领取最高 500 美元的 API 额度，用于构建应用和智能体。Baoyu 报道称，这些额度无需额外付费，ChatGPT Plus 订阅用户不符合条件。 Simon Willison 报道称，订阅符合条件的 100 美元、200 美元或 500 美元套餐的用户，可获得与订阅价格等额的 API 额度，并可将其用于一个 API 组织。

2个来源 · 4位参与者
26

继续看No.04–10

  1. 04

    NVIDIA 发布 Nemotron 3.5 Transcribe，vLLM 从发布首日便提供服务支持

    NVIDIA released Nemotron 3.5 Transcribe, an open-weight speech model, with vLLM announcing launch-day serving support through vLLM and NeMo Speech. NVIDIA described it as a promptable 30B mixture-of-experts model with approximately 4B active parameters that combines transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding in one checkpoint. The Decoder reported that the model follows text instructions to process recordings…

    3个来源
    23↓ 16%
  2. 05

    Google DeepMind 发布 Genie 4 预览版，新增音频、更长模拟时长并计划开放 API 访问

    Google DeepMind has released Genie 4, a model that generates interactive virtual worlds, as a research preview for trusted testers, with consumer and developer access still planned. Google says access is coming soon to Google AI Ultra subscribers through Project Genie and early access partners through the Gemini API and Google AI Studio; TechCrunch reports that rollout is planned in the coming weeks. Google has also opened a Project Genie waitlist. DeepMind says Genie 4 adds…

    5个来源
    16↓ 16%
  3. 06

    Meta AI 的 Muse 模型在 Artificial Analysis 友好度排名中位居榜首

    Artificial Analysis ranked Meta AI’s Muse models highest for friendliness in AA-PersonalityBench, with Muse Spark 1.3 scoring 100, Muse Spark Open 98 and Muse Wonder 95. The benchmark’s Friendly trait measures whether a response conveys warmth and approachability without feeling performative. Alexandr Wang echoed the ranking, calling Muse models “the friendliest of all the models” and adding “personal superintelligence.” The reported benchmark result concerns response friend…

    2个来源
    13↓ 15%
  4. 07

    Google 扩展 Chrome 自动填充功能，通过 Gemini 提供 Gmail 建议

    Google has expanded Chrome’s enhanced autofill to suggest information from users’ Gmail when they fill out online forms, with a rollout underway in the US through Gemini in Chrome on desktop, Android, and iOS. The feature uses Gemini with Personal Intelligence and is available to users who have Personal Intelligence turned on. Suggestions can include flight confirmation numbers, membership IDs, shipment tracking numbers, hotel reservation details, and purchase order numbers.…

    2个来源
    10↓ 16%
  5. 08

    LimX Dynamics 更新 FluxVLA Engine V2，公布在线强化学习结果并扩大机器人评测范围

    LimX Dynamics updated its FluxVLA Engine technical report to V2, adding online reinforcement learning experiments and expanding simulation evaluations and physical robot deployment coverage; reported gains remain specific to the tested tasks and configurations. The open-source framework connects robotics data processing, training, evaluation, inference, and deployment. The updated report covers 14 policies evaluated on LIBERO and reports inference frequency gains of 2.31–9.6…

    2个来源
    9↓ 16%
  6. 09

    修订后的 Artificial Analysis 指数将 Kimi K3.1 Pro 排在第九位，该模型正式发布

    Moonshot AI released Kimi K3.1 Pro through its apps and API, but its weights remain unreleased, and Artificial Analysis’s revised Intelligence Index now places it ninth, behind eight closed models. The Decoder reported that version 5 assigns 60% of the index to private test sets to counter training on known questions. K3.1 Pro scored 46, trailing the leader by 12 points; under version 4.3, it ranked fourth with 55, three points behind Claude Opus 5.5. Ethan Mollick urged caut…

    2个来源
    8↓ 16%
  7. 10

    Anthropic 与六家药物研发机构合作，并启动自主药物发现工作

    Anthropic says its scientists have formed research partnerships with six organizations developing new medicines and have also begun their own drug discovery efforts. The partnerships have been underway for the past few months. According to Baoyu’s account of Anthropic’s announcement, Claude is being used for drug design, experiments and preparation of regulatory submission materials. That account says human clinical trials could begin as early as the end of 2028; it presents…

    2个来源
    6↓ 16%
热度是怎么算的？了解榜单

热度来自参与同一事件的独立账号与机构，重复采集只算一次，并按 24 小时半衰期衰减。它衡量讨论活跃程度，不是报道质量评分。

榜单统计过去 48 小时。趋势只比较持续覆盖的同一组信源；它反映我们的监测范围，不代表全网人数。缺少可比历史时，不展示趋势线。

信源名单只展示可公开阅读的报道来源；讨论参与者还包括只计入热度的账号与机构。同一机构的多个渠道可能合并计数，因此参与者不一定多于信源数。点击事件可查看各方报道与观点。

爆
讨论快速增加
新
首报 6 小时内
发酵中
讨论仍在增加