[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-deepseek-r1-distill-llama-8b":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":139,"fitting_gpus":203,"community":234},101,"deepseek-r1-distill-llama-8b","DeepSeek R1 Distill Llama 8B","deepseek-ai\u002FDeepSeek-R1-Distill-Llama-8B","deepseek-ai","transformer",null,8000000000,"8B",131072,"MIT",326870,872,"published","https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-R1-Distill-Llama-8B","2026-08-12T03:11:11+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},"DeepSeek","deepseek",0.0102,[],[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":114,"GGUF":127},{"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",4096,14.9,23.8,35.7,62.25,"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",16384,64,138.16,207.25,127.25,{"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",8192,2,40.41,60.61,88.25,{"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},7.45,14.37,32,50.62,{"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},127.88,191.82,115.62,{"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},30.55,45.83,76.62,{"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":106,"recommended":110},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":103,"kv_cache_gb":45,"vram_gb":104,"ram_gb":87,"disk_gb":105,"requirement_source":67,"confidence":68},3.86,9.84,45.03,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":103,"kv_cache_gb":72,"vram_gb":107,"ram_gb":108,"disk_gb":109,"requirement_source":67,"confidence":68},122.93,184.4,110.03,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":103,"kv_cache_gb":44,"vram_gb":111,"ram_gb":112,"disk_gb":113,"requirement_source":67,"confidence":68},25.81,38.72,71.03,{"minimum":115,"production":119,"recommended":123},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":116,"kv_cache_gb":45,"vram_gb":117,"ram_gb":87,"disk_gb":118,"requirement_source":67,"confidence":68},3.91,9.9,45.1,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":116,"kv_cache_gb":72,"vram_gb":120,"ram_gb":121,"disk_gb":122,"requirement_source":67,"confidence":68},123,184.5,110.1,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":116,"kv_cache_gb":44,"vram_gb":124,"ram_gb":125,"disk_gb":126,"requirement_source":67,"confidence":68},25.87,38.81,71.1,{"minimum":128,"production":131,"recommended":135},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":34,"kv_cache_gb":45,"vram_gb":129,"ram_gb":87,"disk_gb":130,"requirement_source":67,"confidence":68},10.02,45.25,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":34,"kv_cache_gb":72,"vram_gb":132,"ram_gb":133,"disk_gb":134,"requirement_source":67,"confidence":68},123.13,184.69,110.25,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":34,"kv_cache_gb":44,"vram_gb":136,"ram_gb":137,"disk_gb":138,"requirement_source":67,"confidence":68},26,38.99,71.25,{"recommended":140,"cheapest":158,"performance":168,"min_complexity":185,"cluster_options":199,"no_match":200,"required_vram_gb":111,"candidate_count":201,"usage":202},{"gpu_model":141,"gpu_model_id":79,"gpu_count":40,"total_vram_gb":87,"hour_price":142,"monthly_cost":143,"providers":144,"bandwidth_gbps":147,"fp16_tflops":148,"interconnect_score":149,"parallel_efficiency":40,"score":150,"tier":151,"perf_metric":152,"perf_label":153,"perf_value":154,"workload":10,"cost_metric":155},"RTX 5090",0.3222,235.21,[145,146],"vast","runpod",1792,209,40,0.6828,"cheapest","tok\u002Fs","464.2 tok\u002Fs",464.2,{"effective_value":156,"effective_unit":152,"capacity_factor":157},324.9,0.7,{"gpu_model":159,"gpu_model_id":160,"gpu_count":79,"total_vram_gb":87,"hour_price":161,"monthly_cost":162,"providers":163,"bandwidth_gbps":41,"fp16_tflops":41,"interconnect_score":149,"parallel_efficiency":164,"score":165,"tier":151,"perf_metric":152,"perf_label":166,"perf_value":41,"workload":10,"cost_metric":167},"Tesla V100",17,0.0272,39.71,[145],0.8,0.6316,"0 tok\u002Fs",{"effective_value":41,"effective_unit":152,"capacity_factor":157},{"gpu_model":169,"gpu_model_id":170,"gpu_count":44,"total_vram_gb":171,"hour_price":172,"monthly_cost":173,"providers":174,"bandwidth_gbps":176,"fp16_tflops":177,"interconnect_score":149,"parallel_efficiency":178,"score":179,"tier":180,"perf_metric":152,"perf_label":181,"perf_value":182,"workload":10,"cost_metric":183},"H200 SXM",11,1128,3.5,20440,[175],"lambda",4800,989,0.4,0.2197,"balanced","3979.3 tok\u002Fs",3979.3,{"effective_value":184,"effective_unit":152,"capacity_factor":157},2785.5,{"gpu_model":186,"gpu_model_id":187,"gpu_count":40,"total_vram_gb":188,"hour_price":189,"monthly_cost":190,"providers":191,"bandwidth_gbps":192,"fp16_tflops":193,"interconnect_score":149,"parallel_efficiency":40,"score":194,"tier":151,"perf_metric":152,"perf_label":195,"perf_value":196,"workload":10,"cost_metric":197},"RTX A6000",6,48,0.3001,219.07,[145,175],768,77,0.6343,"199 tok\u002Fs",199,{"effective_value":198,"effective_unit":152,"capacity_factor":157},139.3,[],false,24,"value",[204,208,212,220,227],{"name":186,"vram_gb":188,"cheapest_hour":189,"cost":205,"provider":145,"estimated_tps":196},{"hourly":189,"daily":206,"monthly":190,"yearly":207},7.2,2628.88,{"name":141,"vram_gb":87,"cheapest_hour":142,"cost":209,"provider":145,"estimated_tps":154},{"hourly":142,"daily":210,"monthly":143,"yearly":211},7.73,2822.47,{"name":213,"vram_gb":188,"cheapest_hour":214,"cost":215,"provider":145,"estimated_tps":219},"L40S",0.4669,{"hourly":214,"daily":216,"monthly":217,"yearly":218},11.21,340.84,4090.04,223.8,{"name":221,"vram_gb":188,"cheapest_hour":222,"cost":223,"provider":145,"estimated_tps":41},"RTX PRO 6000 WS",0.8022,{"hourly":222,"daily":224,"monthly":225,"yearly":226},19.25,585.61,7027.27,{"name":228,"vram_gb":188,"cheapest_hour":229,"cost":230,"provider":145,"estimated_tps":41},"RTX PRO 6000 S",1.0131,{"hourly":229,"daily":231,"monthly":232,"yearly":233},24.31,739.56,8874.76,{"posts_count":41,"benchmarks_count":41}]