[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-glm-5.1":3},[4,152],{"id":5,"aliases":6,"types":7,"pricing":10,"capabilities":25,"model_capabilities":31,"model_data":105,"introduction":148,"description":149,"title_focus":150,"keyword_extras":151},"glm-5.1",[5],[8,9],"openai","anthropic",{"input_tokens_tiers":11},{"tiers":12},[13,19],{"max_tokens":14,"pricing":15},32768,{"input_tokens":16,"output_tokens":17,"cache_read_tokens":18,"cache_write_tokens":16},"6","24","1.3",{"max_tokens":20,"pricing":21},0,{"input_tokens":22,"output_tokens":23,"cache_read_tokens":24,"cache_write_tokens":22},"8","28","2",{"context_window":26,"max_output":27,"released_at":28,"modalities":29},200000,131072,"2026-04-07",[30],"text",{"groups":32,"highlights":94,"overall":98,"scenario":101,"summary":103,"weakness":104},[33,71,85],{"items":34,"name":70},[35,41,46,51,55,59,63,67],{"detail":36,"method":37,"name":38,"samples":39,"score":40},"","主\u002F客观混合","智能体协同",7,49,{"detail":36,"method":42,"name":43,"samples":44,"score":45},"客观","代码生成",10,90,{"detail":36,"method":47,"name":48,"samples":49,"score":50},"主观","代码理解与调试",12,85,{"detail":36,"method":47,"name":52,"samples":53,"score":54},"创意写作",17,81,{"detail":36,"method":37,"name":56,"samples":57,"score":58},"信息检索 \u002F RAG",13,92,{"detail":36,"method":47,"name":60,"samples":61,"score":62},"长文本处理",3,95,{"detail":36,"method":42,"name":64,"samples":65,"score":66},"数学与科学推理",14,93,{"detail":36,"method":37,"name":68,"samples":44,"score":69},"多语言能力",88,"专项进阶能力",{"items":72,"name":84},[73,77,80],{"detail":36,"method":37,"name":74,"samples":75,"score":76},"通识知识",15,78,{"detail":36,"method":47,"name":78,"samples":79,"score":66},"文本理解",5,{"detail":36,"method":37,"name":81,"samples":82,"score":83},"逻辑推理",20,70,"通用基础能力",{"items":86,"name":93},[87,90],{"detail":36,"method":47,"name":88,"samples":89,"score":58},"幻觉控制",9,{"detail":36,"method":47,"name":91,"samples":65,"score":92},"指令遵循",82,"安全与合规能力",[95,96,97],{"name":60,"score":62},{"name":78,"score":66},{"name":64,"score":66},{"rating":99,"score":100},"优秀",84,{"fit":100,"name":102},"代码研发 \u002F 内容创作","GLM 5.1在本轮能力评测中综合得分84，整体表现优秀。其长文本处理（95%）和文本理解（93%）表现突出，能够连续理解长篇材料并完成跨段落信息整合，并分析文档语义、提取关键信息并处理复杂问答。对企业知识库团队、AI 应用开发者和研究人员而言，它适合用于资料检索、文档问答、信息抽取和检索增强生成场景，解决从大量资料中找到相关信息，并形成可用于后续决策的回答；数学与科学推理（93%）也为相关任务提供补充。使用时应留意，智能体协同（49%）相对较弱，在需要多步骤自主规划、连续工具调用或跨智能体协作的任务中，建议增加人工复核或专项验证。",{"name":38,"note":36,"score":40},{"bars":106,"metrics":110,"recentAvailability":136,"summary":137},{"degraded":107,"down":109},[108],1,[20],[111,114,117,121,125,129,133],{"label":112,"value":113},"完成率","99%",{"label":115,"value":116},"缓存命中率","动态变化",{"label":118,"unit":119,"value":120},"累计评测耗时","s","8967.2",{"label":122,"unit":123,"value":124},"Token 消耗","tokens","520,646",{"label":126,"unit":127,"value":128},"平均请求次数","次","1.0",{"label":130,"unit":131,"value":132},"TPM","tokens\u002Fmin","3,484",{"label":134,"unit":135,"value":128},"QPM","req\u002Fmin","99.33%",[138,141,144],{"label":139,"unit":119,"value":140},"延迟","38.16",{"label":142,"unit":131,"value":143},"吞吐量","3.48K",{"label":145,"unit":146,"value":147},"可用率","%","99.33","GLM 5.1的主要优势是数学与科学推理、文本理解。在数学求解、科学推理与结构化分析中，它能够承担信息理解、分析和结果生成等关键环节。","GLM 5.1具备长文本处理、文本理解、数学与科学推理能力，可完成长篇材料理解和跨段落信息整合，并支持文档分析、信息抽取和复杂问答。面向企业知识库团队、AI 应用开发者和研究人员，适合资料检索、文档问答、信息抽取和检索增强生成；可通过 SilvaMux 统一 API 一键快速接入。","长文本理解",[60,78],{"id":5,"aliases":153,"types":154,"pricing":156,"capabilities":163,"tier":165,"priority":166,"model_capabilities":167,"model_data":200,"introduction":148,"description":149,"title_focus":150,"keyword_extras":225},[5],[8,9,155],"openai\u002Fresponses\u002Fv1",{"input_tokens_tiers":157},{"tiers":158},[159,161],{"max_tokens":14,"pricing":160},{"input_tokens":16,"output_tokens":17,"cache_read_tokens":18,"cache_write_tokens":16},{"max_tokens":20,"pricing":162},{"input_tokens":22,"output_tokens":23,"cache_read_tokens":24,"cache_write_tokens":22},{"context_window":26,"max_output":27,"released_at":28,"modalities":164},[30],"tune",2,{"groups":168,"highlights":193,"overall":197,"scenario":198,"summary":103,"weakness":199},[169,183,189],{"items":170,"name":70},[171,173,174,176,178,179,181,182],{"detail":36,"method":37,"name":38,"samples":39,"score":172},51,{"detail":36,"method":42,"name":43,"samples":44,"score":45},{"detail":36,"method":37,"name":48,"samples":49,"score":175},77,{"detail":36,"method":47,"name":52,"samples":53,"score":177},83,{"detail":36,"method":37,"name":56,"samples":57,"score":69},{"detail":36,"method":47,"name":60,"samples":61,"score":180},87,{"detail":36,"method":42,"name":64,"samples":65,"score":66},{"detail":36,"method":37,"name":68,"samples":44,"score":45},{"items":184,"name":84},[185,186,187],{"detail":36,"method":37,"name":74,"samples":75,"score":92},{"detail":36,"method":47,"name":78,"samples":79,"score":58},{"detail":36,"method":37,"name":81,"samples":82,"score":188},57,{"items":190,"name":93},[191,192],{"detail":36,"method":47,"name":88,"samples":89,"score":175},{"detail":36,"method":47,"name":91,"samples":65,"score":100},[194,195,196],{"name":64,"score":66},{"name":78,"score":58},{"name":43,"score":45},{"rating":99,"score":54},{"fit":54,"name":102},{"name":38,"note":36,"score":172},{"bars":201,"metrics":204,"recentAvailability":217,"summary":218},{"degraded":202,"down":203},[],[20,108,166],[205,207,208,210,212,213,215],{"label":112,"value":206},"95%",{"label":115,"value":116},{"label":118,"unit":119,"value":209},"19004.2",{"label":122,"unit":123,"value":211},"385,833",{"label":126,"unit":127,"value":128},{"label":130,"unit":131,"value":214},"1,218",{"label":134,"unit":135,"value":216},"0.5","98.66%",[219,221,223],{"label":139,"unit":119,"value":220},"51.26",{"label":142,"unit":131,"value":222},"1.22K",{"label":145,"unit":146,"value":224},"98.66",[60,78]]