[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-glm-legal-harvey-lab-cheapest":3,"topics-all":38,"news-related-27240196-3ac0-4af3-aec7-b0a4d3e7c990":57},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":24,"news_slug":31,"published_at":32,"created_at":33,"modified_at":34,"is_published":35,"publish_type":36,"image_url":14,"view_count":37},"27240196-3ac0-4af3-aec7-b0a4d3e7c990","智谱GLM-Legal法律Agent模型:Harvey榜前十最便宜","智谱Z.ai基于GLM-5.3-Flash发布法律智能体模型GLM-Legal,在律所工作流基准Harvey LAB拿到19.2%全通过率,较底座提升12.5个百分点;Vals独立复测列全榜第7,单测0.15美元,为前十中最便宜。每百万输入token 0.075美元,1M上下文,权重开放。","10 月 12 日,智谱 Z.ai 发布了法律领域智能体模型 GLM-Legal。它不是又一个通用旗舰,而是基于轻量级的 GLM-5.3-Flash 底座,用法律语料和数万个任务环境专门训出来的「专科模型」——目标很直接:把合同审查、证据分析、合规调研这些律师日常重活,做成端到端能交付的 Agent 任务。\n\n## 成绩单:19.2% 全通过率,榜上前十里最便宜\n\n官方评测里,GLM-Legal 在 Harvey LAB(律所真实工作流基准,覆盖 24 个执业领域)拿到 19.2% 的全通过率,比底座 GLM-5.3-Flash 提升 12.5 个百分点;官方同时称其在 LegalBench 和中文基准 PLawBench 上追平闭源前沿模型。\n\n独立评估机构 Vals 的复测给出更具体的坐标:Harvey LAB 19.17%(±3.25),全榜 79 个模型里排第 7,单次测试成本 0.15 美元,是前十名里最便宜的;LegalBench 86.24%,150 个模型里排 14,是 GLM 系列在该基准上的最好成绩;Legal Research Bench 41.83%,78 个模型里排 24。定价为每百万输入 token 0.075 美元、输出 0.25 美元,上下文窗口 1M,权重开放。\n\n## 训练方法:中训打底,强化学习练手,框架一起迭代\n\n这篇发布博客最有信息量的不是跑分,而是「模型与 Agent 框架共同进化」的思路。GLM-Legal 的训练分三层:\n\n- **中训**:扩充高质量法律语料,让模型在密集领域语境里学会法条、解释框架和「事实—规则—结论」的演绎链路;\n- **后训**:在数万个拟真法律环境里做强化学习,模型学会在具体约束下规划分析、推进流程、按交付规格产出文件;\n- **生产反馈**:从业者按「要点是否全覆盖、引用是否站得住、交付物能否直接用」评估产出,问题回流到任务环境、评估细则和框架设计,变成下一轮训练的信号。\n\n博客特别强调,Harness(运行时编排框架)不是外挂工具,而是和模型锁步迭代的一等公民:训练中暴露的边界情况会反推框架改架构,框架改进又把真实执行约束喂回模型。\n\n## 落地:AutoClaw 上的合同、证据与合规\n\n发布同时,Z.ai 展示了桌面 Agent 平台 AutoClaw 上的三类任务。合同审查的对比最能说明差异:底座模型只会建议「考虑缩短通知期」,GLM-Legal 直接给出可执行改法——把便利终止通知期从 60 天压到 30 天、对知识产权\u002F保密\u002F数据安全违约取消补救期允许立即终止、加上最长 90 天的过渡协助,并以修订模式写进 Word 文档。证据分析场景里,它能把「对外报告 60% 完成」与「内部快照约 41%」这类矛盾文件串成证据链,事实与推断分开标注。合规调研场景里,它把新加坡、马来西亚、印尼三地隐私合规拆成带法条引用的优先级清单。\n\n## 所以呢\n\n两个判断。第一,「小底座 + 领域中训 + 数万环境 RL」这条路线,正在把垂直榜的单位成本打下来——前十名里单价最低,说明跑分和账单可以同时赢。第二,19.2% 的全通过率也意味着约八成任务仍会在某个环节掉链子,法律 Agent 离「能独立接案」还有明显距离。但对律所和法务团队,今天就能用的定位是清晰的:一个不会累的初级助理,先把初审和检索的脏活接走。专科模型的战争,才刚开始。\n\n(来源:[Z.ai 官方博客](https:\u002F\u002Fz.ai\u002Fblog\u002Fglm-legal)、[Vals 独立评估](https:\u002F\u002Fwww.vals.ai\u002Fmodels\u002Fzai_glm-legal))","https:\u002F\u002Fz.ai\u002Fblog\u002Fglm-legal","df9ef325-77c5-4e95-9c03-f6cf5b150ef0",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"6ad31a14-c0da-42df-81fd-564281f768db","agentic-ai",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"120fa59a-ff6f-4537-9bf5-f818df636a0e","benchmark",{"id":19,"name":20,"slug":20,"description":14,"color":14},"01598627-1ea6-4b27-a5d8-874971571a71","llm",{"id":22,"name":23,"slug":23,"description":14,"color":14},"7e89b5cc-57db-4f37-bc6d-28919a73931c","model-release",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"f2ac547d-144e-4a70-92bf-0384dc297bcc","en","Z.ai GLM-Legal: cheapest model in Harvey LAB top ten","Z.ai's GLM-Legal scores 19.2% all-pass on Harvey LAB, ranked 7\u002F79 by Vals at $0.15\u002Ftest — cheapest in the top ten. 1M context, open weights.","On October 12, Z.ai released GLM-Legal, an agentic legal intelligence model. It is not another general-purpose flagship: built on the lightweight GLM-5.3-Flash base and trained with legal corpora and tens of thousands of task environments, it is a \"specialist\" aimed at turning contract review, evidence analysis and compliance research into end-to-end agent deliverables.\n\n## The Score: 19.2% All-Pass, Cheapest in the Top Ten\n\nIn the official evaluation, GLM-Legal scored a 19.2% all-pass rate on Harvey LAB, a benchmark of real law-firm workflows across 24 practice areas — a 12.5 percentage-point improvement over its base model. Z.ai also reports parity with frontier closed-source models on LegalBench and the Chinese-language PLawBench.\n\nIndependent evaluator Vals puts it in concrete coordinates: 19.17% (±3.25) on Harvey LAB, ranking 7th of 79 models, at $0.15 per test — the cheapest model in the top ten. On LegalBench it scores 86.24%, ranking 14th of 150 and marking the best GLM-family result on that benchmark; on Legal Research Bench it reaches 41.83%, 24th of 78. Pricing is $0.075 per million input tokens and $0.25 per million output tokens, with a 1M context window and open weights.\n\n## Training: Mid-Training Foundations, RL Practice, Co-Evolving Harness\n\nThe most informative part of the release blog is not the scores but the idea of co-evolving the model with its agent harness. Training has three layers:\n\n- **Mid-training** expands high-quality legal corpora so the model learns statutes, interpretive frameworks and the fact-rule-conclusion deductive chain in dense domain context;\n- **Post-training** applies reinforcement learning across tens of thousands of realistic legal environments, teaching the model to plan analysis, advance workflows and produce deliverables under concrete constraints;\n- **Production feedback** has practitioners grade outputs against criteria like issue coverage, citation soundness and immediate usability, and the findings flow back into task environments, rubrics and harness design.\n\nThe blog stresses that the Harness — the runtime orchestration framework — is not an external tool but a first-class citizen iterating in lockstep with the model: edge cases found during training expose harness bottlenecks, while harness refinements feed realistic execution constraints back into the model.\n\n## In Practice: Contracts, Evidence and Compliance on AutoClaw\n\nAlongside the release, Z.ai demonstrated three task types on AutoClaw, its desktop agent platform. The contract-review contrast is the most telling: where the base model suggests \"consider shortening the notice period,\" GLM-Legal proposes executable edits — cutting the convenience-termination notice from 60 to 30 days, removing the cure period for IP, confidentiality or data-security breaches so the client can terminate immediately, and adding up to 90 days of transition assistance — written into the Word document with tracked changes. In evidence analysis it chains contradictory documents such as an external \"60% complete\" report against an internal ~41% snapshot, keeping facts and inferences distinct. In compliance research it turns privacy obligations across Singapore, Malaysia and Indonesia into a prioritized checklist with statutory references.\n\n## So What\n\nTwo observations. First, the recipe of \"small base + domain mid-training + RL across tens of thousands of environments\" is driving down the unit cost of vertical leaderboards — being the cheapest in the top ten shows scores and bills can be won at the same time. Second, a 19.2% all-pass rate means roughly four in five tasks still break somewhere along the chain, and legal agents remain far from taking cases independently. For law firms and in-house teams, though, the usable positioning today is clear: a tireless junior associate who takes over first-pass review and research. The specialist-model war has just begun.\n\n(Sources: [Z.ai blog](https:\u002F\u002Fz.ai\u002Fblog\u002Fglm-legal), [Vals evaluation](https:\u002F\u002Fwww.vals.ai\u002Fmodels\u002Fzai_glm-legal))","glm-legal-harvey-lab-cheapest","2026-10-13T15:20:00Z","2026-10-13T15:09:16.353473Z","2026-10-13T15:09:16.353486Z",true,"agent",27,[39,48],{"slug":40,"tag_slug":40,"title_zh":41,"title_en":42,"intro_zh":43,"intro_en":44,"id":45,"is_active":35,"created_at":46,"modified_at":47},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":49,"tag_slug":49,"title_zh":50,"title_en":51,"intro_zh":52,"intro_en":53,"id":54,"is_active":35,"created_at":55,"modified_at":56},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":58},[59,64,69,74,78,83],{"id":60,"title":61,"news_slug":62,"published_at":63},"c4ec4625-4a84-4c24-88f0-0ef1beb4f19e","Grok 4.6 发布:61 分追平 GPT-5.6 Sol,把长程 Agent 的 token 账单砍到四分之一","grok-4-6-agentic-cost-frontier","2026-08-14T19:00:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"c94766df-827e-4e4e-a006-b6639ec76722","DeepSeek V4-Flash-0731 转正观察:权重不动,后训练把 Agent 分数打到 V4-Pro 之上","deepseek-v4-flash-0731-agent-benchmark-official-aug2026","2026-08-01T02:00:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"3cc63477-1334-497d-80cb-90850c019101","DeepSeek-V4-Flash 转正:不靠换架构,只做后训练重新发力 Agent","deepseek-v4-flash-official-post-training-agent-0731","2026-07-31T08:00:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":73},"fbdcfd41-bb54-487a-8319-9f35dfc82be5","DeepSeek-V4-Flash转正：这次升级不靠换架构","deepseek-v4-flash-0731-agent-update",{"id":79,"title":80,"news_slug":81,"published_at":82},"859242dd-0035-41e6-8db9-4583989048b7","AgentSWE基准:让编码智能体造智能体,无人及格","agentswe-coding-agent-benchmark","2026-10-13T19:08:00+00:00",{"id":84,"title":85,"news_slug":86,"published_at":87},"9c8731f9-e170-4855-93a2-b1705abb61c3","Muse Wonder发布:Meta把宝押在个人智能体上","meta-muse-wonder-agentic-model","2026-10-12T23:10:00+00:00"]