[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-glm-47-flash":3},{"id":4,"slug":5,"name":6,"hf_id":7,"organization":8,"architecture":9,"architecture_config":10,"parameters":11,"parameters_label":12,"context_length":13,"license":14,"downloads":15,"likes":16,"status":17,"description":10,"huggingface_url":18,"created_at":19,"seo":20,"category":23,"family":26,"score":29,"variants":30,"quantizations":31,"requirements":58,"gpu_recommendations":143,"fitting_gpus":214,"community":249},147,"glm-47-flash","GLM-4.7-Flash","zai-org\u002FGLM-4.7-Flash","zai-org","moe_transformer",null,30000000000,"30B",131072,"MIT",1865359,1806,"published","https:\u002F\u002Fhuggingface.co\u002Fzai-org\u002FGLM-4.7-Flash","2026-08-12T03:11:13+00:00",{"title":6,"description":21,"image":10,"robots":22,"canonical_url":10},"AI 模型部署决策引擎——开源 AI 的 GPU 需求、云端价格与成本分析。","index, follow",{"name":24,"slug":25},"LLM","llm",{"name":27,"slug":28},"GLM","glm",0.0853,[],[32,37,42,47,51,55],{"format":33,"bits":34,"size_factor":35,"quality_loss":36},"AWQ",4,0.263,0.03,{"format":38,"bits":39,"size_factor":40,"quality_loss":41},"FP16",16,1,0,{"format":43,"bits":44,"size_factor":45,"quality_loss":46},"FP8",8,0.5,0.01,{"format":48,"bits":34,"size_factor":49,"quality_loss":50},"GGUF",0.25,0.05,{"format":52,"bits":34,"size_factor":53,"quality_loss":54},"GPTQ",0.275,0.04,{"format":56,"bits":44,"size_factor":45,"quality_loss":57},"INT8",0.02,{"FP16":59,"FP8":83,"INT8":97,"AWQ":101,"GPTQ":115,"GGUF":129},{"minimum":60,"production":69,"recommended":76},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":63,"kv_cache_gb":45,"vram_gb":64,"ram_gb":65,"disk_gb":66,"requirement_source":67,"confidence":68},"minimum",4096,55.88,75.64,113.46,126.17,"estimated",0.95,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":63,"kv_cache_gb":72,"vram_gb":73,"ram_gb":74,"disk_gb":75,"requirement_source":67,"confidence":68},"production",16384,64,194.71,292.07,191.17,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":63,"kv_cache_gb":44,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":67,"confidence":68},"recommended",8192,2,94.6,141.9,152.17,{"minimum":84,"production":89,"recommended":93},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":85,"kv_cache_gb":45,"vram_gb":86,"ram_gb":87,"disk_gb":88,"requirement_source":67,"confidence":68},27.94,40.29,60.44,82.59,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":85,"kv_cache_gb":72,"vram_gb":90,"ram_gb":91,"disk_gb":92,"requirement_source":67,"confidence":68},156.16,234.24,147.59,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":85,"kv_cache_gb":44,"vram_gb":94,"ram_gb":95,"disk_gb":96,"requirement_source":67,"confidence":68},57.65,86.48,108.59,{"minimum":98,"production":99,"recommended":100},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":85,"kv_cache_gb":45,"vram_gb":86,"ram_gb":87,"disk_gb":88,"requirement_source":67,"confidence":68},{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":85,"kv_cache_gb":72,"vram_gb":90,"ram_gb":91,"disk_gb":92,"requirement_source":67,"confidence":68},{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":85,"kv_cache_gb":44,"vram_gb":94,"ram_gb":95,"disk_gb":96,"requirement_source":67,"confidence":68},{"minimum":102,"production":107,"recommended":111},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":103,"kv_cache_gb":45,"vram_gb":104,"ram_gb":105,"disk_gb":106,"requirement_source":67,"confidence":68},14.49,23.28,34.93,61.61,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":103,"kv_cache_gb":72,"vram_gb":108,"ram_gb":109,"disk_gb":110,"requirement_source":67,"confidence":68},137.6,206.4,126.61,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":103,"kv_cache_gb":44,"vram_gb":112,"ram_gb":113,"disk_gb":114,"requirement_source":67,"confidence":68},39.87,59.8,87.61,{"minimum":116,"production":121,"recommended":125},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":117,"kv_cache_gb":45,"vram_gb":118,"ram_gb":119,"disk_gb":120,"requirement_source":67,"confidence":68},14.67,23.51,35.26,61.88,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":117,"kv_cache_gb":72,"vram_gb":122,"ram_gb":123,"disk_gb":124,"requirement_source":67,"confidence":68},137.84,206.76,126.88,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":117,"kv_cache_gb":44,"vram_gb":126,"ram_gb":127,"disk_gb":128,"requirement_source":67,"confidence":68},40.1,60.15,87.88,{"minimum":130,"production":135,"recommended":139},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":131,"kv_cache_gb":45,"vram_gb":132,"ram_gb":133,"disk_gb":134,"requirement_source":67,"confidence":68},15.02,23.95,35.92,62.43,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":131,"kv_cache_gb":72,"vram_gb":136,"ram_gb":137,"disk_gb":138,"requirement_source":67,"confidence":68},138.32,207.49,127.43,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":131,"kv_cache_gb":44,"vram_gb":140,"ram_gb":141,"disk_gb":142,"requirement_source":67,"confidence":68},40.56,60.84,88.43,{"recommended":144,"cheapest":166,"performance":182,"min_complexity":197,"cluster_options":210,"no_match":211,"required_vram_gb":112,"candidate_count":212,"usage":213},{"gpu_model":145,"gpu_model_id":146,"gpu_count":40,"total_vram_gb":147,"hour_price":148,"monthly_cost":149,"providers":150,"bandwidth_gbps":155,"fp16_tflops":156,"interconnect_score":157,"parallel_efficiency":40,"score":158,"tier":159,"perf_metric":160,"perf_label":161,"perf_value":162,"workload":10,"cost_metric":163},"H100 SXM",10,80,1.6059,1172.31,[151,152,153,154],"vast","tensordock","runpod","lambda",3350,66.9,40,0.6224,"balanced","tok\u002Fs","231.2 tok\u002Fs",231.2,{"effective_value":164,"effective_unit":160,"capacity_factor":165},161.8,0.7,{"gpu_model":167,"gpu_model_id":168,"gpu_count":34,"total_vram_gb":169,"hour_price":170,"monthly_cost":171,"providers":172,"bandwidth_gbps":173,"fp16_tflops":174,"interconnect_score":157,"parallel_efficiency":175,"score":176,"tier":177,"perf_metric":160,"perf_label":178,"perf_value":179,"workload":10,"cost_metric":180},"RTX 5070",28,48,0.0101,29.49,[151],672,61,0.6,0.5894,"cheapest","111.3 tok\u002Fs",111.3,{"effective_value":181,"effective_unit":160,"capacity_factor":165},77.9,{"gpu_model":183,"gpu_model_id":184,"gpu_count":44,"total_vram_gb":185,"hour_price":186,"monthly_cost":187,"providers":188,"bandwidth_gbps":189,"fp16_tflops":190,"interconnect_score":157,"parallel_efficiency":191,"score":192,"tier":159,"perf_metric":160,"perf_label":193,"perf_value":194,"workload":10,"cost_metric":195},"H200 SXM",11,1128,3.5,20440,[154],4800,989,0.4,0.2197,"1060 tok\u002Fs",1060,{"effective_value":196,"effective_unit":160,"capacity_factor":165},742,{"gpu_model":198,"gpu_model_id":199,"gpu_count":40,"total_vram_gb":169,"hour_price":200,"monthly_cost":201,"providers":202,"bandwidth_gbps":203,"fp16_tflops":204,"interconnect_score":157,"parallel_efficiency":40,"score":205,"tier":177,"perf_metric":160,"perf_label":206,"perf_value":207,"workload":10,"cost_metric":208},"RTX A6000",6,0.2874,209.8,[151,154],768,77,0.6696,"53 tok\u002Fs",53,{"effective_value":209,"effective_unit":160,"capacity_factor":165},37.1,[],false,22,"value",[215,219,227,234,241],{"name":198,"vram_gb":169,"cheapest_hour":200,"cost":216,"provider":151,"estimated_tps":207},{"hourly":200,"daily":217,"monthly":201,"yearly":218},6.9,2517.62,{"name":220,"vram_gb":169,"cheapest_hour":221,"cost":222,"provider":151,"estimated_tps":226},"L40S",0.4004,{"hourly":221,"daily":223,"monthly":224,"yearly":225},9.61,292.29,3507.5,59.6,{"name":228,"vram_gb":169,"cheapest_hour":229,"cost":230,"provider":151,"estimated_tps":41},"RTX PRO 6000 WS",0.6689,{"hourly":229,"daily":231,"monthly":232,"yearly":233},16.05,488.3,5859.56,{"name":235,"vram_gb":169,"cheapest_hour":236,"cost":237,"provider":151,"estimated_tps":41},"RTX PRO 6000 S",0.7339,{"hourly":236,"daily":238,"monthly":239,"yearly":240},17.61,535.75,6428.96,{"name":242,"vram_gb":147,"cheapest_hour":243,"cost":244,"provider":151,"estimated_tps":248},"A100 80GB",1.2,{"hourly":243,"daily":245,"monthly":246,"yearly":247},28.8,876,10512,140.7,{"posts_count":41,"benchmarks_count":41}]