月之暗面发布 Kimi K3.1 Pro，Terminal-Bench 4.0 得分 65.7%
Moonshot AI’s first update, the new Kimi K3.1 Pro
月之暗面发布 Kimi K3.1 Pro，这是其首个更新版本，在 Terminal-Bench 4.0 上得分 65.7%，比 Claude Opus 5.5 低 0.7 分、比 GPT-6 Astra 高 7.5 分，且平均每任务成本低于两者。该模型为 3T 级开放权重 MoE 模型，总参数 2.8T、激活参数 104B，支持 1M token 上下文和多模态；在带工具的幻觉压力测试中，错误声明率从 K3 的 23.4% 降至 8.3%。K3.1 Pro 与速度快 2-3 倍的 Highspeed 版本已在 Kimi 应用和 API 上线，完整权重计划于 11 月 4 日发布。
推荐理由
精选AI 评分 77
原文把 Kimi K3.1 Pro 与 Claude Opus 5.5、GPT-6 Astra 在多项基准和成本上并列比较，并给出能力不均衡的具体位置。
主题
查看原帖隆重推出 Kimi K3.1 Pro：我们迄今能力最强的 3T 级开放权重模型。 🔹 1M token 上下文，总参数 2.8T / 激活参数 104B（MoE），多模态。 🔹 全项目级编程：可在 Unreal、Unity 和 Godot 中开发游戏。 🔹 擅长在 Blender 和 CAD 中进行参数化 3D 建模。 🔹 在金融工作上更可靠：建模准确，幻觉率降低 65%。 🔹 在编程、智能体和知识基准测试中，与最强的专有模型不相上下。 Kimi K3.1 Pro 和 K3.1 Pro Highspeed（速度提升 2–3 倍）现已在 https://t.co/85jA71X9gw、Kimi Work、Kimi Code 和 Kimi API 上线。 开放权重将于 2026 年 11 月 4 日发布。 🔗 API：https://t.co/XCrgjXAqMw 🔗 技术博客：https://t.co/5iPP0MBej6
查看原文
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
正文
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.