腾讯混元开源交互式视频世界模型 Hy-Nomas
#模型发布#开源生态#视频
#模型发布#开源生态#视频
#模型发布#开源生态#评测/基准
Available now: NVIDIA Nemotron 3.5 Bolt E4B 4GB ⚡️ Our newest small open model is built for local assistants, interactive characters and robots that can carry on a conversation and use tools to get things done. Watch Reachy Mini chat and move, with Bolt handling conversation and NVIDIA Cosmos handling vision. Both models run locally on Jetson Orin Nano 8GB. Learn more: https://t.co/B8ODaWMn0O
#模型发布#端侧#开源生态
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model. 🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal. 🔹 Full-project coding: games in Unreal, Unity & Godot. 🔹 Excels at parametric 3D modeling in Blender & CAD. 🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%. 🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks. Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on https://t.co/85jA71X9gw, Kimi Work, Kimi Code, and the Kimi API. Open weights by Nov 4, 2026. 🔗 API: https://t.co/XCrgjXAqMw 🔗 Tech blog: https://t.co/5iPP0MBej6
#评测/基准#编码#开源生态
#模型发布#推理#Agent
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model. 🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal. 🔹 Full-project coding: games in Unreal, Unity & Godot. 🔹 Excels at parametric 3D modeling in Blender & CAD. 🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%. 🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks. Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on https://t.co/85jA71X9gw, Kimi Work, Kimi Code, and the Kimi API. Open weights by Nov 4, 2026. 🔗 API: https://t.co/XCrgjXAqMw 🔗 Tech blog: https://t.co/5iPP0MBej6
#模型发布#Agent#推理
Moonshot AI’s first update, the new Kimi K3.1 Pro > Kimi K3.1 Pro scores 65.7% on Terminal-Bench 4.0, 0.7 points behind Claude Opus 5.5 and 7.5 ahead of GPT-6 Astra, at a lower average cost per task than either rival’s top runs. > Moonshot saw exploit skill climb as post-training scaled, so it removed cyber-related data from mid-training and RL; on its internal xdev-bench-hard, pass@3 fell from 90.5% to 81.0%. > The filtered model still solves 81.0% of those exploitation tasks against K3’s 33.3%, which Moonshot reads as general coding gains carrying into cyber work, while a third-party run of GPT-5.6 Sol solved 100%. > On Moonshot’s internal hallucination stress test with tools, false claims fell from 23.4% for K3 to 8.3%, still above GPT-6 Astra’s 5.5% but well below Claude Opus 5.5’s 15.0%. > The biggest jump is on Moonshot’s simulated business-operation test, where scores rose from 8.9 to 37, and K3.1 Pro also tops OfficeQA Pro v2 at 64.4 against Opus 5.5’s 62.2. > Gains are uneven: K3.1 Pro scores 38.1 on Terminal-Bench Science against GPT-6 Astra’s 68.1 and Opus 5.5’s 63.3, while leading Agents’ Last Exam at 43.2 and Humanity’s Last Exam with tools at 70.0. > K3.1 Pro and a Highspeed variant running 2-3x faster are live in Kimi’s apps and API, with full weights due by November 4 for independent reruns.
#模型发布#安全/对齐#开源生态
#模型发布#开源生态#评测/基准
Kimi K3.1 Pro launches at 4th on the Artificial Analysis Intelligence Index scoring 55, making large gains in agentic performance; it will be the leading open-weights model once its weights are published @Kimi_Moonshot's new reasoning model gains 11 points over Kimi K3 (max, 44). It trails only Claude Opus 5.5 (max with fallback, 58), Claude Sonnet 5.5 (max with fallback, 56), and Muse Wonder Full (max, 56) on the Intelligence Index. Once the weights for K3.1 Pro are released it will be the highest-scoring open weights model and the gap between the leading open weights and proprietary models on the Intelligence Index will narrow from 12 points to 3. Key results: ➤ Large gains in agentic performance: Kimi K3.1 Pro scores 3rd on GDPval-AA with 1,846 Elo, up 312 Elo from Kimi K3, behind only Claude Opus 5.5 (1,866 Elo) and Muse Wonder Full (1,857 Elo). Kimi K3.1 Pro also achieves the highest score of any model on Terminal-Bench 4.0. K3.1 Pro scores 67%, up from 13% for Kimi K3 and ahead of Claude Sonnet 5.5 (64%). AutomationBench-AA, which tests agentic SaaS workflows, rises from 58% to 73%. ➤ Gains in other evals are more moderate: Humanity's Last Exam rises from 47% to 52%, CritPt from 23% to 25%, and SciCode from 59% to 61%. AA-Omniscience Index rises from 20 to 30, as the hallucination rate falls from 53% to 44% and accuracy rises from 48% to 51%. ➤ Cost per Task roughly equal to Kimi K3 at $1.98 per Task: K3.1 Pro uses 55K output tokens per Intelligence Index task, 14% more than Kimi K3 at 48K, while its first-party token prices are 33% lower. Additional model details: ➤ Context window: 1M tokens. ➤ Size: 2.8T total, 104B active parameters ➤ Input modalities: Text and image ➤ Weights and license: Weights currently not published, Moonshot is expected to release the weights soon after launch ➤ Pricing: $2/$10/$0.20 per 1M input/output/cache hit tokens, 33% lower than Kimi K3 on all three rates. Congratulations to the team at @Kimi_Moonshot on the release!
#大佬观点#评测/基准#模型发布
蚂蚁数科发布决策模型LingDT-Decision-1.0,提供4B和27B两个版本,面向客服、风控、金融等场景的高频决策任务,已上线DTMaaS平台限时免费体验。该模型支持原生多选与按序决策(DAG),在23项开源评测基准中27B版本综合评分第一,超过Microsoft-Decision-1、Jev 1.13.0等模型。蚂蚁数科测试显示,数据标注场景相比DeepSeek-V4-flash Token消耗降低约97%、处理时效提升约10倍,智能客服场景相比Qwen-3.7-flash Token消耗降低90%、风险识别准确率由76%提升至89%。
#模型发布#Agent#推理
Moonshot AI 于 10 月 24 日发布 Kimi K3.1 Pro,这是 K3.1 系列的首个模型,定位为 3T 级开源权重模型,支持 100 万 token 上下文窗口,同时推出速度提升 2 至 3 倍的 Kimi K3.1 Pro Highspeed。官方称 K3.1 Pro 在 Terminal-Bench 4.0 上得分 65.7%,低于 Claude Opus 5.5 的 66.4%、高于 GPT-6 Astra 的 58.2%;内部幻觉压力测试中,其声明级幻觉率从 K3 的 23.4% 降至 8.3%。两款模型已上线 Kimi.ai、移动端、Kimi Work 桌面端、Kimi Code 和 Kimi API,完整模型权重将于 11 月 4 日发布。
#模型发布#开源生态#Agent
Block 内部 AI 研究实验室于 10 月 23 日发布全双工语音到语音模型 Dessa,称其完全基于人类对话语音从零训练,可同时听和说,并能原生感知情绪、讽刺和笑声。该模型推理速度快于实时,可在 MacBook Pro M4 Max 或更高配置上本地运行,团队用单台 NVIDIA GB200 NVL72 完成训练。Block 已公开演示录音和公开等待名单,并向少量 AI 安全研究者开放早期访问。
#模型发布#语音#开源生态
Jack Dorsey 旗下 Block 发布 AI 模型 Dessa,称其完全基于人类对话从零训练,是真正的语音到语音模型。官方展示了多段对话样例,包括一段与 OpenAI GPT-Live-1 语音模型的对比,Dessa 在其中显得更自然。Dessa 可在 MacBook Pro M4 Max 或更高配置上本地运行,团队表示将向一小批 AI 安全研究者开放早期访问。
#模型发布#语音#开源生态
月之暗面发布推理模型 Kimi K3.1 Pro,在 Artificial Analysis 智能指数上得分 55,排名第 4,比 Kimi K3 的 44 分提升 11 分,仅次于 Claude Opus 5.5(58)、Claude Sonnet 5.5(56)和 Muse Wonder Full(56)。其智能体能力提升明显,GDPval-AA 得分 1,846 Elo 排第 3,Terminal-Bench 4.0 得分 67%(Kimi K3 为 13%),AutomationBench-AA 从 58% 升至 73%。模型上下文窗口 1M tokens,总参数 2.8T、激活参数 104B,支持文本和图像输入,定价为每 1M 输入/输出/缓存命中 token 2/10/0.20 美元,三项均比 Kimi K3 低 33%;权重尚未公布,发布后将成为得分最高的开源权重模型,开源与闭源领先模型的差距将从 12 分缩小到 3 分。
#模型发布#推理#Agent
Moonshot AI 发布 K3.1 系列首个模型 Kimi K3.1 Pro,采用 MoE 架构,总参数 2.8T、激活 104B,支持 1M token 上下文窗口。该模型已在 Kimi.ai、Kimi Work、Kimi Code 和 OpenAI 兼容的 Kimi API 上线,API 定价为每百万 token 输入 2 美元、输出 10 美元,低于 K3 的 3 美元和 15 美元。Moonshot 自报 Terminal-Bench 4.0 得分 65.7%,SWE-Marathon 51.7 领先其图表中所有模型,但 Terminal-Bench Science 0.1 仅 38.1,落后 GPT-6 Astra 的 68.1;开放权重预计 2026 年 11 月发布。
#模型发布#开源生态#Agent
月之暗面推出旗舰模型Kimi K3.1 Pro,总参数2.8万亿、采用MoE架构、每次激活1040亿参数,支持100万Token上下文窗口,并计划在11月4日前开放完整模型权重。该模型在Terminal-Bench 4.0编程智能体测试中得分65.7%,高于GPT-6 Astra的58.2%,略低于Claude Opus 5.5的66.4%;BenchCAD基准借助工具得分92.7%。标准版API输出价68元/百万Token,高速版速度约为标准版2至3倍、价格为两倍,已接入Kimi官网、App、Kimi Work、Kimi Code和API。
#模型发布#Agent#编码
月之暗面发布 Kimi K3.1 Pro,这是 K3.1 系列首个模型,也是其迄今最强的 3T 级开源权重模型,上下文窗口为 100 万 token,同时推出速度提升 2-3 倍的 Kimi K3.1 Pro Highspeed。该模型已在 Kimi.com、Kimi Work、Kimi Code 和 Kimi API 上线,完整模型权重将于 11 月 4 日发布。在 Terminal-Bench 4.0 上,K3.1 Pro 得分 65.7%,高于 GPT-6 Astra 的 58.2%,略低于 Claude Opus 5.5 的 66.4%;在内部基准 Kimi Code Bench 3.0 Long-horizon 上得分为 61.2%,高于 K3 的 55.95%。

#模型发布#开源生态#Agent
Google 发布新前沿模型 Gemini 4 Argon,称其在编码、数学和 3D 世界构建等复杂任务上表现突出,并在多领域基准上超过其他前沿模型。官方展示了 5 个社区项目,包括实时映射野火与地震波动的 3D 地球、带实时数据的涡扇发动机模拟、由单张图片生成的 3D 马车,以及多人夺旗浏览器游戏。该模型上周已大范围开放。

#产品更新#Google#多模态
Kimi K3.1 Pro 今日上线 Kimi API,目前尚未开放权重,月之暗面承诺 2026 年 11 月 4 日前发布完整模型权重。定价为输入 $2.00/百万 token、输出 $10/百万 token,与 GPT-6.1 Sol 相同,为 Claude Opus 5.5 的一半;另提供速度快 2-3 倍、价格翻倍的 Highspeed 版本。Artificial Analysis 给 Kimi 打出 55 分,高于 Claude Fable 5.1、GPT-6 Astra 和 Gemini 4 Argon 的 53 分,仅次于 Claude Opus 5.5 的 58 分。Simon Willison 通过 Moonshot API 在 low、medium、max 三档推理强度下实测了生成 SVG 的效果。
#模型发布#DeepSeek#推理
#产品更新#视频#多模态