[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-qwen2-vl-2b":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":28,"variants":29,"quantizations":30,"requirements":57,"gpu_recommendations":131,"fitting_gpus":191,"community":219},339,"qwen2-vl-2b","Qwen2-VL 2B","Qwen\u002FQwen2-VL-2B-Instruct","Qwen","transformer",null,2000000000,"2B",32768,"MIT",2470447,518,"published","https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen2-VL-2B-Instruct","2026-08-12T03:11:23+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},"OCR \u002F Vision","ocr-vision",{"name":8,"slug":27},"qwen",0.0096,[],[31,36,41,46,50,54],{"format":32,"bits":33,"size_factor":34,"quality_loss":35},"AWQ",4,0.263,0.03,{"format":37,"bits":38,"size_factor":39,"quality_loss":40},"FP16",16,1,0,{"format":42,"bits":43,"size_factor":44,"quality_loss":45},"FP8",8,0.5,0.01,{"format":47,"bits":33,"size_factor":48,"quality_loss":49},"GGUF",0.25,0.05,{"format":51,"bits":33,"size_factor":52,"quality_loss":53},"GPTQ",0.275,0.04,{"format":55,"bits":43,"size_factor":44,"quality_loss":56},"INT8",0.02,{"FP16":58,"FP8":80,"INT8":92,"AWQ":96,"GPTQ":108,"GGUF":120},{"minimum":59,"production":68,"recommended":75},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":62,"kv_cache_gb":48,"vram_gb":63,"ram_gb":64,"disk_gb":65,"requirement_source":66,"confidence":67},"minimum",2048,3.73,8.29,32,44.81,"estimated",0.95,{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":62,"kv_cache_gb":43,"vram_gb":72,"ram_gb":73,"disk_gb":74,"requirement_source":66,"confidence":67},"production",8192,2,26.74,40.11,109.81,{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":62,"kv_cache_gb":39,"vram_gb":78,"ram_gb":64,"disk_gb":79,"requirement_source":66,"confidence":67},"recommended",4096,10.68,70.81,{"minimum":81,"production":85,"recommended":89},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":82,"kv_cache_gb":48,"vram_gb":83,"ram_gb":64,"disk_gb":84,"requirement_source":66,"confidence":67},1.86,5.93,41.91,{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":82,"kv_cache_gb":43,"vram_gb":86,"ram_gb":87,"disk_gb":88,"requirement_source":66,"confidence":67},24.17,36.26,106.91,{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":82,"kv_cache_gb":39,"vram_gb":90,"ram_gb":64,"disk_gb":91,"requirement_source":66,"confidence":67},8.21,67.91,{"minimum":93,"production":94,"recommended":95},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":82,"kv_cache_gb":48,"vram_gb":83,"ram_gb":64,"disk_gb":84,"requirement_source":66,"confidence":67},{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":82,"kv_cache_gb":43,"vram_gb":86,"ram_gb":87,"disk_gb":88,"requirement_source":66,"confidence":67},{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":82,"kv_cache_gb":39,"vram_gb":90,"ram_gb":64,"disk_gb":91,"requirement_source":66,"confidence":67},{"minimum":97,"production":101,"recommended":105},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":98,"kv_cache_gb":48,"vram_gb":99,"ram_gb":64,"disk_gb":100,"requirement_source":66,"confidence":67},0.97,4.8,40.51,{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":98,"kv_cache_gb":43,"vram_gb":102,"ram_gb":103,"disk_gb":104,"requirement_source":66,"confidence":67},22.93,34.4,105.51,{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":98,"kv_cache_gb":39,"vram_gb":106,"ram_gb":64,"disk_gb":107,"requirement_source":66,"confidence":67},7.03,66.51,{"minimum":109,"production":113,"recommended":117},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":110,"kv_cache_gb":48,"vram_gb":111,"ram_gb":64,"disk_gb":112,"requirement_source":66,"confidence":67},0.98,4.81,40.53,{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":110,"kv_cache_gb":43,"vram_gb":114,"ram_gb":115,"disk_gb":116,"requirement_source":66,"confidence":67},22.95,34.42,105.53,{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":110,"kv_cache_gb":39,"vram_gb":118,"ram_gb":64,"disk_gb":119,"requirement_source":66,"confidence":67},7.04,66.53,{"minimum":121,"production":124,"recommended":128},{"scenario":60,"context_length":61,"batch_size":39,"weight_gb":39,"kv_cache_gb":48,"vram_gb":122,"ram_gb":64,"disk_gb":123,"requirement_source":66,"confidence":67},4.84,40.56,{"scenario":69,"context_length":70,"batch_size":71,"weight_gb":39,"kv_cache_gb":43,"vram_gb":125,"ram_gb":126,"disk_gb":127,"requirement_source":66,"confidence":67},22.98,34.47,105.56,{"scenario":76,"context_length":77,"batch_size":39,"weight_gb":39,"kv_cache_gb":39,"vram_gb":129,"ram_gb":64,"disk_gb":130,"requirement_source":66,"confidence":67},7.07,66.56,{"recommended":132,"cheapest":150,"performance":159,"min_complexity":175,"cluster_options":187,"no_match":188,"required_vram_gb":106,"candidate_count":189,"usage":190},{"gpu_model":133,"gpu_model_id":134,"gpu_count":39,"total_vram_gb":38,"hour_price":135,"monthly_cost":136,"providers":137,"bandwidth_gbps":139,"fp16_tflops":140,"interconnect_score":141,"parallel_efficiency":39,"score":142,"tier":143,"perf_metric":144,"perf_label":145,"perf_value":146,"workload":10,"cost_metric":147},"RTX 5080",23,0.1078,78.69,[138],"vast",960,113,40,0.6124,"cheapest","tok\u002Fs","989.7 tok\u002Fs",989.7,{"effective_value":148,"effective_unit":144,"capacity_factor":149},692.8,0.7,{"gpu_model":151,"gpu_model_id":152,"gpu_count":39,"total_vram_gb":38,"hour_price":153,"monthly_cost":154,"providers":155,"bandwidth_gbps":40,"fp16_tflops":40,"interconnect_score":141,"parallel_efficiency":39,"score":156,"tier":143,"perf_metric":144,"perf_label":157,"perf_value":40,"workload":10,"cost_metric":158},"Tesla V100",17,0.0272,19.86,[138],0.6068,"0 tok\u002Fs",{"effective_value":40,"effective_unit":144,"capacity_factor":149},{"gpu_model":160,"gpu_model_id":71,"gpu_count":43,"total_vram_gb":161,"hour_price":162,"monthly_cost":163,"providers":164,"bandwidth_gbps":166,"fp16_tflops":167,"interconnect_score":141,"parallel_efficiency":168,"score":169,"tier":170,"perf_metric":144,"perf_label":171,"perf_value":172,"workload":10,"cost_metric":173},"RTX 5090",256,0.3222,1881.65,[138,165],"runpod",1792,209,0.4,0.3618,"balanced","5911.8 tok\u002Fs",5911.8,{"effective_value":174,"effective_unit":144,"capacity_factor":149},4138.3,{"gpu_model":176,"gpu_model_id":177,"gpu_count":39,"total_vram_gb":43,"hour_price":178,"monthly_cost":179,"providers":180,"bandwidth_gbps":181,"fp16_tflops":177,"interconnect_score":141,"parallel_efficiency":39,"score":182,"tier":143,"perf_metric":144,"perf_label":183,"perf_value":184,"workload":10,"cost_metric":185},"RTX 5060 Ti",24,0.0554,40.44,[138],448,0.658,"461.9 tok\u002Fs",461.9,{"effective_value":186,"effective_unit":144,"capacity_factor":149},323.3,[],false,18,"value",[192,196,200,208,211],{"name":151,"vram_gb":38,"cheapest_hour":153,"cost":193,"provider":138,"estimated_tps":40},{"hourly":153,"daily":194,"monthly":154,"yearly":195},0.65,238.27,{"name":176,"vram_gb":43,"cheapest_hour":178,"cost":197,"provider":138,"estimated_tps":184},{"hourly":178,"daily":198,"monthly":179,"yearly":199},1.33,485.3,{"name":201,"vram_gb":177,"cheapest_hour":202,"cost":203,"provider":138,"estimated_tps":207},"RTX 3090",0.0678,{"hourly":202,"daily":204,"monthly":205,"yearly":206},1.63,49.49,593.93,964.9,{"name":209,"vram_gb":38,"cheapest_hour":202,"cost":210,"provider":138,"estimated_tps":40},"RTX 4070S Ti",{"hourly":202,"daily":204,"monthly":205,"yearly":206},{"name":212,"vram_gb":213,"cheapest_hour":214,"cost":215,"provider":138,"estimated_tps":148},"RTX 5070",12,0.0804,{"hourly":214,"daily":216,"monthly":217,"yearly":218},1.93,58.69,704.3,{"posts_count":40,"benchmarks_count":40}]