[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-image-service-v2-gemini-2.5-pro":3,"text-service-v2-gemini-2.5-pro":208},[4],{"id":5,"aliases":6,"types":7,"pricing_group":9,"pricing":10,"display_name":5,"capabilities":26,"model_capabilities":31,"model_capabilities_zh":105,"introduction":154,"introduction_zh":155,"description":156,"description_zh":157,"title_focus":158,"title_focus_zh":159,"keyword_extras":160,"keyword_extras_zh":163,"model_data":166},"gemini-2.5-pro",[5],[8],"openai","google\u002Fgemini",{"input_tokens_tiers":11},{"tiers":12},[13,20],{"max_tokens":14,"pricing":15},200000,{"input_tokens":16,"output_tokens":17,"cache_read_tokens":18,"cache_write_tokens":19},"1.25","10","0.125","0",{"max_tokens":21,"pricing":22},0,{"input_tokens":23,"output_tokens":24,"cache_read_tokens":25,"cache_write_tokens":19},"2.5","15","0.25",{"context_window":27,"max_output":28,"modalities":29},1048576,65536,[30],"text",{"groups":32,"highlights":94,"overall":98,"scenario":101,"summary":103,"weakness":104},[33,71,85],{"items":34,"name":70},[35,41,45,49,54,58,62,66],{"detail":36,"method":37,"name":38,"samples":39,"score":40},"","Mixed","Agent Collaboration",18,56,{"detail":36,"method":42,"name":43,"samples":44,"score":40},"Objective","Code Generation",12,{"detail":36,"method":37,"name":46,"samples":47,"score":48},"Code Comprehension & Debugging",6,63,{"detail":36,"method":50,"name":51,"samples":52,"score":53},"Subjective","Creative Writing",22,65,{"detail":36,"method":37,"name":55,"samples":56,"score":57},"Information Retrieval \u002F RAG",9,66,{"detail":36,"method":50,"name":59,"samples":60,"score":61},"Long-Context Handling",16,76,{"detail":36,"method":37,"name":63,"samples":64,"score":65},"Math & Science Reasoning",14,85,{"detail":36,"method":50,"name":67,"samples":68,"score":69},"Multilingual Capability",11,86,"Advanced Capabilities",{"items":72,"name":84},[73,77,81],{"detail":36,"method":37,"name":74,"samples":75,"score":76},"General Knowledge",10,82,{"detail":36,"method":50,"name":78,"samples":79,"score":80},"Text Comprehension",23,77,{"detail":36,"method":37,"name":82,"samples":60,"score":83},"Logical Reasoning",69,"General Foundations",{"items":86,"name":93},[87,91],{"detail":36,"method":50,"name":88,"samples":89,"score":90},"Hallucination Control",15,75,{"detail":36,"method":50,"name":92,"samples":39,"score":90},"Instruction Following","Safety & Compliance",[95,96,97],{"name":67,"score":69},{"name":63,"score":65},{"name":74,"score":76},{"rating":99,"score":100},"良好",72,{"fit":80,"name":102},"Translation & Localization \u002F Content Moderation & Risk Control \u002F Customer Support","Gemini 2.5 Pro scored 72 overall, performing well — a multimodal thinker built for long context. It stands out in multilingual capability (86%), math and science reasoning (85%) and general knowledge (82%), handling cross-language material and rigorous analysis over long documents, which serves translation and content moderation well. Comparatively, agent collaboration (56%) is weak; tool-orchestration workflows are a stretch. Long-document multilingual analysis is its sweet spot.",{"name":38,"note":36,"score":40},{"groups":106,"highlights":145,"overall":149,"scenario":150,"summary":152,"weakness":153},[107,129,138],{"items":108,"name":128},[109,112,115,117,120,122,124,126],{"detail":36,"method":110,"name":111,"samples":39,"score":40},"主\u002F客观混合","智能体协同",{"detail":36,"method":113,"name":114,"samples":44,"score":40},"客观","代码生成",{"detail":36,"method":110,"name":116,"samples":47,"score":48},"代码理解与调试",{"detail":36,"method":118,"name":119,"samples":52,"score":53},"主观","创意写作",{"detail":36,"method":110,"name":121,"samples":56,"score":57},"信息检索 \u002F RAG",{"detail":36,"method":118,"name":123,"samples":60,"score":61},"长文本处理",{"detail":36,"method":110,"name":125,"samples":64,"score":65},"数学与科学推理",{"detail":36,"method":118,"name":127,"samples":68,"score":69},"多语言能力","专项进阶能力",{"items":130,"name":137},[131,133,135],{"detail":36,"method":110,"name":132,"samples":75,"score":76},"通识知识",{"detail":36,"method":118,"name":134,"samples":79,"score":80},"文本理解",{"detail":36,"method":110,"name":136,"samples":60,"score":83},"逻辑推理","通用基础能力",{"items":139,"name":144},[140,142],{"detail":36,"method":118,"name":141,"samples":89,"score":90},"幻觉控制",{"detail":36,"method":118,"name":143,"samples":39,"score":90},"指令遵循","安全与合规能力",[146,147,148],{"name":127,"score":69},{"name":125,"score":65},{"name":132,"score":76},{"rating":99,"score":100},{"fit":80,"name":151},"翻译与本地化 \u002F 内容审核与风控 \u002F 智能客服","Gemini 2.5 Pro 综合得分 72，整体表现良好，是一款面向长上下文的多模态思考模型。多语言能力（86%）、数学与科学推理（85%）与通识知识（82%）表现突出，能在长文档中完成跨语言材料的严谨分析，在翻译本地化与内容审核场景中表现稳定。相对而言，智能体协同（56%）偏弱，工具编排类工作流有些勉强。长文档多语言分析是它的最佳场景。",{"name":111,"note":36,"score":40},"Gemini 2.5 Pro is Google's multimodal thinking model with long context and native tool calls, strong in multilingual, mathematical and analytical work — suited to research, translation and content review.","Gemini 2.5 Pro是Google的多模态思考模型，支持长上下文与原生工具调用，擅长多语言、数学推理和复杂资料分析，适合研究、翻译与内容审核场景。","Gemini 2.5 Pro brings multimodal understanding, long-context analysis and reasoning to complex document processing and multilingual tasks. Integrate in one click via the SilvaMux unified API.","Gemini 2.5 Pro具备多模态理解、长上下文分析与推理能力，适合复杂资料处理和多语言任务。可通过 SilvaMux 统一 API 一键快速接入。","Multimodal Reasoning Model","多模态推理模型",[161,162],"long context","multimodal",[164,165],"长上下文","多模态",{"summary":167,"recentAvailability":180,"observedAt":36,"bars":181,"metrics":186},[168,172,176],{"label":169,"value":170,"unit":171},"modelsDetail.performance.latency","30.38","s",{"label":173,"value":174,"unit":175},"modelsDetail.performance.throughput","50.28K","tokens\u002Fmin",{"label":177,"value":178,"unit":179},"modelsDetail.performance.availability","99.2","%","99.2%",{"degraded":182,"down":184},[183],3,[185],30,[187,190,193,197,201,204],{"label":188,"value":189},"modelsDetail.performance.completionRate","95%",{"label":191,"value":192},"modelsDetail.performance.cacheHitRate","80%",{"label":194,"value":195,"unit":196},"modelsDetail.performance.tokenUsage","2,332,058,123","tokens",{"label":198,"value":199,"unit":200},"modelsDetail.performance.avgRequests","47","modelsDetail.performance.requestsUnit",{"label":202,"value":203,"unit":175},"modelsDetail.performance.tpm","50,283",{"label":205,"value":206,"unit":207},"modelsDetail.performance.qpm","930","req\u002Fmin",{"ttft":209,"tps":212,"e2e":213,"requests":216,"successfulRequests":217,"successRate":218,"availability":219},{"p50":210,"p95":211},2496,2500,{"p50":83,"p95":90},{"p50":214,"p95":215},3818,6714,163044,160301,"98.32",[220,220,220,220,220,220,220,220,220,220,220,220,220,220,220,220,221,220,220,220,220,220,220,220],"up","degraded"]