[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-codellama-13b":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":210,"community":245},75,"codellama-13b","CodeLlama 13B","codellama\u002FCodeLlama-13b-hf","codellama","transformer",null,13000000000,"13B",16384,"MIT",2941,118,"published","https:\u002F\u002Fhuggingface.co\u002Fcodellama\u002FCodeLlama-13b-hf","2026-08-12T03:11:10+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},"Coding AI","coding",{"name":27,"slug":28},"Llama","llama",0.0013,[],[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":82,"INT8":96,"AWQ":100,"GPTQ":113,"GGUF":126},{"minimum":60,"production":69,"recommended":75},{"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",4096,24.21,35.58,53.37,76.77,"estimated",0.95,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":63,"kv_cache_gb":71,"vram_gb":72,"ram_gb":73,"disk_gb":74,"requirement_source":67,"confidence":68},"production",64,151.02,226.52,141.77,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":63,"kv_cache_gb":44,"vram_gb":79,"ram_gb":80,"disk_gb":81,"requirement_source":67,"confidence":68},"recommended",8192,2,52.72,79.09,102.77,{"minimum":83,"production":88,"recommended":92},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":84,"kv_cache_gb":45,"vram_gb":85,"ram_gb":86,"disk_gb":87,"requirement_source":67,"confidence":68},12.11,20.27,32,57.89,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":84,"kv_cache_gb":71,"vram_gb":89,"ram_gb":90,"disk_gb":91,"requirement_source":67,"confidence":68},134.31,201.46,122.89,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":84,"kv_cache_gb":44,"vram_gb":93,"ram_gb":94,"disk_gb":95,"requirement_source":67,"confidence":68},36.71,55.07,83.89,{"minimum":97,"production":98,"recommended":99},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":84,"kv_cache_gb":45,"vram_gb":85,"ram_gb":86,"disk_gb":87,"requirement_source":67,"confidence":68},{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":84,"kv_cache_gb":71,"vram_gb":89,"ram_gb":90,"disk_gb":91,"requirement_source":67,"confidence":68},{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":84,"kv_cache_gb":44,"vram_gb":93,"ram_gb":94,"disk_gb":95,"requirement_source":67,"confidence":68},{"minimum":101,"production":105,"recommended":109},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":102,"kv_cache_gb":45,"vram_gb":103,"ram_gb":86,"disk_gb":104,"requirement_source":67,"confidence":68},6.28,12.89,48.8,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":102,"kv_cache_gb":71,"vram_gb":106,"ram_gb":107,"disk_gb":108,"requirement_source":67,"confidence":68},126.27,189.4,113.8,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":102,"kv_cache_gb":44,"vram_gb":110,"ram_gb":111,"disk_gb":112,"requirement_source":67,"confidence":68},29.01,43.51,74.8,{"minimum":114,"production":118,"recommended":122},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":115,"kv_cache_gb":45,"vram_gb":116,"ram_gb":86,"disk_gb":117,"requirement_source":67,"confidence":68},6.36,12.99,48.92,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":115,"kv_cache_gb":71,"vram_gb":119,"ram_gb":120,"disk_gb":121,"requirement_source":67,"confidence":68},126.37,189.56,113.92,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":115,"kv_cache_gb":44,"vram_gb":123,"ram_gb":124,"disk_gb":125,"requirement_source":67,"confidence":68},29.11,43.66,74.92,{"minimum":127,"production":131,"recommended":135},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":128,"kv_cache_gb":45,"vram_gb":129,"ram_gb":86,"disk_gb":130,"requirement_source":67,"confidence":68},6.51,13.18,49.15,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":128,"kv_cache_gb":71,"vram_gb":132,"ram_gb":133,"disk_gb":134,"requirement_source":67,"confidence":68},126.58,189.87,114.15,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":128,"kv_cache_gb":44,"vram_gb":136,"ram_gb":137,"disk_gb":138,"requirement_source":67,"confidence":68},29.31,43.96,75.15,{"recommended":140,"cheapest":162,"performance":178,"min_complexity":193,"cluster_options":206,"no_match":207,"required_vram_gb":110,"candidate_count":208,"usage":209},{"gpu_model":141,"gpu_model_id":142,"gpu_count":40,"total_vram_gb":143,"hour_price":144,"monthly_cost":145,"providers":146,"bandwidth_gbps":151,"fp16_tflops":152,"interconnect_score":153,"parallel_efficiency":40,"score":154,"tier":155,"perf_metric":156,"perf_label":157,"perf_value":158,"workload":10,"cost_metric":159},"H100 SXM",10,80,1.6059,1172.31,[147,148,149,150],"vast","tensordock","runpod","lambda",3350,66.9,40,0.5862,"balanced","tok\u002Fs","533.4 tok\u002Fs",533.4,{"effective_value":160,"effective_unit":156,"capacity_factor":161},373.4,0.7,{"gpu_model":163,"gpu_model_id":164,"gpu_count":34,"total_vram_gb":165,"hour_price":166,"monthly_cost":167,"providers":168,"bandwidth_gbps":169,"fp16_tflops":170,"interconnect_score":153,"parallel_efficiency":171,"score":172,"tier":173,"perf_metric":156,"perf_label":174,"perf_value":175,"workload":10,"cost_metric":176},"RTX 5070",28,48,0.0101,29.49,[147],672,61,0.6,0.5721,"cheapest","256.8 tok\u002Fs",256.8,{"effective_value":177,"effective_unit":156,"capacity_factor":161},179.8,{"gpu_model":179,"gpu_model_id":180,"gpu_count":44,"total_vram_gb":181,"hour_price":182,"monthly_cost":183,"providers":184,"bandwidth_gbps":185,"fp16_tflops":186,"interconnect_score":153,"parallel_efficiency":187,"score":188,"tier":155,"perf_metric":156,"perf_label":189,"perf_value":190,"workload":10,"cost_metric":191},"H200 SXM",11,1128,3.5,20440,[150],4800,989,0.4,0.2197,"2445.9 tok\u002Fs",2445.9,{"effective_value":192,"effective_unit":156,"capacity_factor":161},1712.1,{"gpu_model":194,"gpu_model_id":195,"gpu_count":40,"total_vram_gb":165,"hour_price":196,"monthly_cost":197,"providers":198,"bandwidth_gbps":199,"fp16_tflops":200,"interconnect_score":153,"parallel_efficiency":40,"score":201,"tier":173,"perf_metric":156,"perf_label":202,"perf_value":203,"workload":10,"cost_metric":204},"RTX A6000",6,0.2874,209.8,[147,150],768,77,0.6523,"122.3 tok\u002Fs",122.3,{"effective_value":205,"effective_unit":156,"capacity_factor":161},85.6,[],false,23,"value",[211,215,223,231,238],{"name":194,"vram_gb":165,"cheapest_hour":196,"cost":212,"provider":147,"estimated_tps":203},{"hourly":196,"daily":213,"monthly":197,"yearly":214},6.9,2517.62,{"name":216,"vram_gb":86,"cheapest_hour":217,"cost":218,"provider":147,"estimated_tps":222},"RTX 5090",0.337,{"hourly":217,"daily":219,"monthly":220,"yearly":221},8.09,246.01,2952.12,285.4,{"name":224,"vram_gb":165,"cheapest_hour":225,"cost":226,"provider":147,"estimated_tps":230},"L40S",0.4004,{"hourly":225,"daily":227,"monthly":228,"yearly":229},9.61,292.29,3507.5,137.6,{"name":232,"vram_gb":165,"cheapest_hour":233,"cost":234,"provider":147,"estimated_tps":41},"RTX PRO 6000 WS",0.6689,{"hourly":233,"daily":235,"monthly":236,"yearly":237},16.05,488.3,5859.56,{"name":239,"vram_gb":165,"cheapest_hour":240,"cost":241,"provider":147,"estimated_tps":41},"RTX PRO 6000 S",0.7339,{"hourly":240,"daily":242,"monthly":243,"yearly":244},17.61,535.75,6428.96,{"posts_count":41,"benchmarks_count":41}]