[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-phind-codellama-34b":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":142,"fitting_gpus":213,"community":248},296,"phind-codellama-34b","Phind-CodeLlama 34B","Phind\u002FPhind-CodeLlama-34B-v2","Phind","transformer",null,34000000000,"34B",16384,"MIT",500,834,"published","https:\u002F\u002Fhuggingface.co\u002FPhind\u002FPhind-CodeLlama-34B-v2","2026-08-12T03:11:21+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",0.0092,[],[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":114,"GGUF":128},{"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",4096,63.33,85.06,127.59,137.79,"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",64,205,307.49,202.79,{"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",8192,2,104.45,156.68,163.79,{"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},31.66,45.01,67.51,88.4,{"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},161.3,241.95,153.4,{"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},62.58,93.87,114.4,{"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":106,"recommended":110},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":102,"kv_cache_gb":45,"vram_gb":103,"ram_gb":104,"disk_gb":105,"requirement_source":67,"confidence":68},16.43,25.73,38.59,64.62,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":102,"kv_cache_gb":71,"vram_gb":107,"ram_gb":108,"disk_gb":109,"requirement_source":67,"confidence":68},140.27,210.4,129.62,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":102,"kv_cache_gb":44,"vram_gb":111,"ram_gb":112,"disk_gb":113,"requirement_source":67,"confidence":68},42.42,63.64,90.62,{"minimum":115,"production":120,"recommended":124},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":116,"kv_cache_gb":45,"vram_gb":117,"ram_gb":118,"disk_gb":119,"requirement_source":67,"confidence":68},16.62,25.98,38.97,64.93,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":116,"kv_cache_gb":71,"vram_gb":121,"ram_gb":122,"disk_gb":123,"requirement_source":67,"confidence":68},140.54,210.81,129.93,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":116,"kv_cache_gb":44,"vram_gb":125,"ram_gb":126,"disk_gb":127,"requirement_source":67,"confidence":68},42.69,64.03,90.93,{"minimum":129,"production":134,"recommended":138},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":130,"kv_cache_gb":45,"vram_gb":131,"ram_gb":132,"disk_gb":133,"requirement_source":67,"confidence":68},17.02,26.48,39.72,65.55,{"scenario":70,"context_length":13,"batch_size":34,"weight_gb":130,"kv_cache_gb":71,"vram_gb":135,"ram_gb":136,"disk_gb":137,"requirement_source":67,"confidence":68},141.09,211.63,130.55,{"scenario":76,"context_length":77,"batch_size":78,"weight_gb":130,"kv_cache_gb":44,"vram_gb":139,"ram_gb":140,"disk_gb":141,"requirement_source":67,"confidence":68},43.21,64.81,91.55,{"recommended":143,"cheapest":165,"performance":181,"min_complexity":196,"cluster_options":209,"no_match":210,"required_vram_gb":111,"candidate_count":211,"usage":212},{"gpu_model":144,"gpu_model_id":145,"gpu_count":40,"total_vram_gb":146,"hour_price":147,"monthly_cost":148,"providers":149,"bandwidth_gbps":154,"fp16_tflops":155,"interconnect_score":156,"parallel_efficiency":40,"score":157,"tier":158,"perf_metric":159,"perf_label":160,"perf_value":161,"workload":10,"cost_metric":162},"H100 SXM",10,80,1.6059,1172.31,[150,151,152,153],"vast","tensordock","runpod","lambda",3350,66.9,40,0.6309,"balanced","tok\u002Fs","203.9 tok\u002Fs",203.9,{"effective_value":163,"effective_unit":159,"capacity_factor":164},142.7,0.7,{"gpu_model":166,"gpu_model_id":167,"gpu_count":34,"total_vram_gb":168,"hour_price":169,"monthly_cost":170,"providers":171,"bandwidth_gbps":172,"fp16_tflops":173,"interconnect_score":156,"parallel_efficiency":174,"score":175,"tier":176,"perf_metric":159,"perf_label":177,"perf_value":178,"workload":10,"cost_metric":179},"RTX 5070",28,48,0.0101,29.49,[150],672,61,0.6,0.5752,"cheapest","98.2 tok\u002Fs",98.2,{"effective_value":180,"effective_unit":159,"capacity_factor":164},68.7,{"gpu_model":182,"gpu_model_id":183,"gpu_count":44,"total_vram_gb":184,"hour_price":185,"monthly_cost":186,"providers":187,"bandwidth_gbps":188,"fp16_tflops":189,"interconnect_score":156,"parallel_efficiency":190,"score":191,"tier":158,"perf_metric":159,"perf_label":192,"perf_value":193,"workload":10,"cost_metric":194},"H200 SXM",11,1128,3.5,20440,[153],4800,989,0.4,0.2197,"934.9 tok\u002Fs",934.9,{"effective_value":195,"effective_unit":159,"capacity_factor":164},654.4,{"gpu_model":197,"gpu_model_id":198,"gpu_count":40,"total_vram_gb":168,"hour_price":199,"monthly_cost":200,"providers":201,"bandwidth_gbps":202,"fp16_tflops":203,"interconnect_score":156,"parallel_efficiency":40,"score":204,"tier":176,"perf_metric":159,"perf_label":205,"perf_value":206,"workload":10,"cost_metric":207},"RTX A6000",6,0.2874,209.8,[150,153],768,77,0.6554,"46.7 tok\u002Fs",46.7,{"effective_value":208,"effective_unit":159,"capacity_factor":164},32.7,[],false,22,"value",[214,218,226,233,240],{"name":197,"vram_gb":168,"cheapest_hour":199,"cost":215,"provider":150,"estimated_tps":206},{"hourly":199,"daily":216,"monthly":200,"yearly":217},6.9,2517.62,{"name":219,"vram_gb":168,"cheapest_hour":220,"cost":221,"provider":150,"estimated_tps":225},"L40S",0.4004,{"hourly":220,"daily":222,"monthly":223,"yearly":224},9.61,292.29,3507.5,52.6,{"name":227,"vram_gb":168,"cheapest_hour":228,"cost":229,"provider":150,"estimated_tps":41},"RTX PRO 6000 WS",0.6689,{"hourly":228,"daily":230,"monthly":231,"yearly":232},16.05,488.3,5859.56,{"name":234,"vram_gb":168,"cheapest_hour":235,"cost":236,"provider":150,"estimated_tps":41},"RTX PRO 6000 S",0.7339,{"hourly":235,"daily":237,"monthly":238,"yearly":239},17.61,535.75,6428.96,{"name":241,"vram_gb":146,"cheapest_hour":242,"cost":243,"provider":150,"estimated_tps":247},"A100 80GB",1.2,{"hourly":242,"daily":244,"monthly":245,"yearly":246},28.8,876,10512,124.1,{"posts_count":41,"benchmarks_count":41}]