[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-phi-35-mini-38b":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":18,"huggingface_url":19,"created_at":20,"seo":21,"category":23,"family":26,"score":29,"variants":30,"quantizations":31,"requirements":58,"gpu_recommendations":139,"fitting_gpus":203,"community":239},25,"phi-35-mini-38b","Phi-3.5 Mini 3.8B","microsoft\u002FPhi-3.5-mini-instruct","microsoft","transformer",null,3820000000,"3.8B",131072,"MIT",610000,1200,"published","3.8B mini coding-capable model, runs anywhere.","https:\u002F\u002Fhuggingface.co\u002Fmicrosoft\u002FPhi-3.5-mini-instruct","2026-08-12T02:22:24+00:00",{"title":6,"description":18,"image":10,"robots":22,"canonical_url":10},"index, follow",{"name":24,"slug":25},"LLM","llm",{"name":27,"slug":28},"Phi","phi",0.0142,[],[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":96,"AWQ":100,"GPTQ":113,"GGUF":126},{"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",4096,7.12,13.95,32,50.1,"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",16384,64,127.42,191.13,115.1,{"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",8192,2,30.11,45.16,76.1,{"minimum":84,"production":88,"recommended":92},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":85,"kv_cache_gb":45,"vram_gb":86,"ram_gb":65,"disk_gb":87,"requirement_source":67,"confidence":68},3.56,9.45,44.55,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":85,"kv_cache_gb":72,"vram_gb":89,"ram_gb":90,"disk_gb":91,"requirement_source":67,"confidence":68},122.51,183.76,109.55,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":85,"kv_cache_gb":44,"vram_gb":93,"ram_gb":94,"disk_gb":95,"requirement_source":67,"confidence":68},25.4,38.11,70.55,{"minimum":97,"production":98,"recommended":99},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":85,"kv_cache_gb":45,"vram_gb":86,"ram_gb":65,"disk_gb":87,"requirement_source":67,"confidence":68},{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":85,"kv_cache_gb":72,"vram_gb":89,"ram_gb":90,"disk_gb":91,"requirement_source":67,"confidence":68},{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":85,"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":65,"disk_gb":104,"requirement_source":67,"confidence":68},1.85,7.28,41.88,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":102,"kv_cache_gb":72,"vram_gb":106,"ram_gb":107,"disk_gb":108,"requirement_source":67,"confidence":68},120.15,180.22,106.88,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":102,"kv_cache_gb":44,"vram_gb":110,"ram_gb":111,"disk_gb":112,"requirement_source":67,"confidence":68},23.14,34.71,67.88,{"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":65,"disk_gb":117,"requirement_source":67,"confidence":68},1.87,7.31,41.91,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":115,"kv_cache_gb":72,"vram_gb":119,"ram_gb":120,"disk_gb":121,"requirement_source":67,"confidence":68},120.18,180.27,106.91,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":115,"kv_cache_gb":44,"vram_gb":123,"ram_gb":124,"disk_gb":125,"requirement_source":67,"confidence":68},23.17,34.76,67.91,{"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":65,"disk_gb":130,"requirement_source":67,"confidence":68},1.91,7.37,41.98,{"scenario":70,"context_length":71,"batch_size":34,"weight_gb":128,"kv_cache_gb":72,"vram_gb":132,"ram_gb":133,"disk_gb":134,"requirement_source":67,"confidence":68},120.24,180.36,106.98,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":128,"kv_cache_gb":44,"vram_gb":136,"ram_gb":137,"disk_gb":138,"requirement_source":67,"confidence":68},23.23,34.84,67.98,{"recommended":140,"cheapest":158,"performance":168,"min_complexity":185,"cluster_options":200,"no_match":201,"required_vram_gb":110,"candidate_count":4,"usage":202},{"gpu_model":141,"gpu_model_id":79,"gpu_count":40,"total_vram_gb":65,"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",1792,209,40,0.6907,"cheapest","tok\u002Fs","968.6 tok\u002Fs",968.6,{"effective_value":156,"effective_unit":152,"capacity_factor":157},678,0.7,{"gpu_model":159,"gpu_model_id":160,"gpu_count":79,"total_vram_gb":65,"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],0.8,0.6395,"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","8302.7 tok\u002Fs",8302.7,{"effective_value":184,"effective_unit":152,"capacity_factor":157},5811.9,{"gpu_model":186,"gpu_model_id":187,"gpu_count":40,"total_vram_gb":188,"hour_price":189,"monthly_cost":190,"providers":191,"bandwidth_gbps":193,"fp16_tflops":194,"interconnect_score":149,"parallel_efficiency":40,"score":195,"tier":151,"perf_metric":152,"perf_label":196,"perf_value":197,"workload":10,"cost_metric":198},"RTX 3090",3,24,0.0678,49.49,[145,192],"tensordock",936,71,0.6423,"505.9 tok\u002Fs",505.9,{"effective_value":199,"effective_unit":152,"capacity_factor":157},354.1,[],false,"value",[204,208,216,223,231],{"name":186,"vram_gb":188,"cheapest_hour":189,"cost":205,"provider":145,"estimated_tps":197},{"hourly":189,"daily":206,"monthly":190,"yearly":207},1.63,593.93,{"name":209,"vram_gb":188,"cheapest_hour":210,"cost":211,"provider":145,"estimated_tps":215},"RTX 4090",0.1344,{"hourly":210,"daily":212,"monthly":213,"yearly":214},3.23,98.11,1177.34,544.9,{"name":217,"vram_gb":188,"cheapest_hour":218,"cost":219,"provider":145,"estimated_tps":215},"RTX 4090D",0.1602,{"hourly":218,"daily":220,"monthly":221,"yearly":222},3.84,116.95,1403.35,{"name":224,"vram_gb":188,"cheapest_hour":225,"cost":226,"provider":145,"estimated_tps":230},"L4",0.2015,{"hourly":225,"daily":227,"monthly":228,"yearly":229},4.84,147.1,1765.14,162.2,{"name":232,"vram_gb":188,"cheapest_hour":233,"cost":234,"provider":145,"estimated_tps":238},"A10",0.2409,{"hourly":233,"daily":235,"monthly":236,"yearly":237},5.78,175.86,2110.28,324.3,{"posts_count":41,"benchmarks_count":41}]