[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-yi-vl-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":141,"fitting_gpus":212,"community":248},407,"yi-vl-34b","Yi-VL 34B","01-ai\u002FYi-VL-34B","01-ai","transformer",null,34000000000,"34B",4096,"MIT",221,265,"published","https:\u002F\u002Fhuggingface.co\u002F01-ai\u002FYi-VL-34B","2026-08-12T03:11:27+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":27,"slug":28},"Yi","yi",0.0029,[],[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":81,"INT8":95,"AWQ":99,"GPTQ":113,"GGUF":127},{"minimum":60,"production":69,"recommended":76},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":63,"kv_cache_gb":49,"vram_gb":64,"ram_gb":65,"disk_gb":66,"requirement_source":67,"confidence":68},"minimum",2048,63.33,83.69,125.53,137.79,"estimated",0.95,{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":63,"kv_cache_gb":44,"vram_gb":73,"ram_gb":74,"disk_gb":75,"requirement_source":67,"confidence":68},"production",8192,2,109,163.49,202.79,{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":63,"kv_cache_gb":40,"vram_gb":78,"ram_gb":79,"disk_gb":80,"requirement_source":67,"confidence":68},"recommended",89.5,134.26,163.79,{"minimum":82,"production":87,"recommended":91},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":83,"kv_cache_gb":49,"vram_gb":84,"ram_gb":85,"disk_gb":86,"requirement_source":67,"confidence":68},31.66,43.63,65.45,88.4,{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":83,"kv_cache_gb":44,"vram_gb":88,"ram_gb":89,"disk_gb":90,"requirement_source":67,"confidence":68},65.3,97.95,153.4,{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":83,"kv_cache_gb":40,"vram_gb":92,"ram_gb":93,"disk_gb":94,"requirement_source":67,"confidence":68},47.63,71.44,114.4,{"minimum":96,"production":97,"recommended":98},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":83,"kv_cache_gb":49,"vram_gb":84,"ram_gb":85,"disk_gb":86,"requirement_source":67,"confidence":68},{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":83,"kv_cache_gb":44,"vram_gb":88,"ram_gb":89,"disk_gb":90,"requirement_source":67,"confidence":68},{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":83,"kv_cache_gb":40,"vram_gb":92,"ram_gb":93,"disk_gb":94,"requirement_source":67,"confidence":68},{"minimum":100,"production":105,"recommended":109},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":101,"kv_cache_gb":49,"vram_gb":102,"ram_gb":103,"disk_gb":104,"requirement_source":67,"confidence":68},16.43,24.35,36.53,64.62,{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":101,"kv_cache_gb":44,"vram_gb":106,"ram_gb":107,"disk_gb":108,"requirement_source":67,"confidence":68},44.27,66.4,129.62,{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":101,"kv_cache_gb":40,"vram_gb":110,"ram_gb":111,"disk_gb":112,"requirement_source":67,"confidence":68},27.47,41.21,90.62,{"minimum":114,"production":119,"recommended":123},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":115,"kv_cache_gb":49,"vram_gb":116,"ram_gb":117,"disk_gb":118,"requirement_source":67,"confidence":68},16.62,24.6,36.91,64.93,{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":115,"kv_cache_gb":44,"vram_gb":120,"ram_gb":121,"disk_gb":122,"requirement_source":67,"confidence":68},44.54,66.81,129.93,{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":115,"kv_cache_gb":40,"vram_gb":124,"ram_gb":125,"disk_gb":126,"requirement_source":67,"confidence":68},27.74,41.6,90.93,{"minimum":128,"production":133,"recommended":137},{"scenario":61,"context_length":62,"batch_size":40,"weight_gb":129,"kv_cache_gb":49,"vram_gb":130,"ram_gb":131,"disk_gb":132,"requirement_source":67,"confidence":68},17.02,25.11,37.66,65.55,{"scenario":70,"context_length":71,"batch_size":72,"weight_gb":129,"kv_cache_gb":44,"vram_gb":134,"ram_gb":135,"disk_gb":136,"requirement_source":67,"confidence":68},45.09,67.63,130.55,{"scenario":77,"context_length":13,"batch_size":40,"weight_gb":129,"kv_cache_gb":40,"vram_gb":138,"ram_gb":139,"disk_gb":140,"requirement_source":67,"confidence":68},28.26,42.39,91.55,{"recommended":142,"cheapest":164,"performance":180,"min_complexity":195,"cluster_options":208,"no_match":209,"required_vram_gb":110,"candidate_count":210,"usage":211},{"gpu_model":143,"gpu_model_id":144,"gpu_count":40,"total_vram_gb":145,"hour_price":146,"monthly_cost":147,"providers":148,"bandwidth_gbps":153,"fp16_tflops":154,"interconnect_score":155,"parallel_efficiency":40,"score":156,"tier":157,"perf_metric":158,"perf_label":159,"perf_value":160,"workload":10,"cost_metric":161},"H100 SXM",10,80,1.6059,1172.31,[149,150,151,152],"vast","tensordock","runpod","lambda",3350,66.9,40,0.5811,"balanced","tok\u002Fs","203.9 tok\u002Fs",203.9,{"effective_value":162,"effective_unit":158,"capacity_factor":163},142.7,0.7,{"gpu_model":165,"gpu_model_id":166,"gpu_count":34,"total_vram_gb":167,"hour_price":168,"monthly_cost":169,"providers":170,"bandwidth_gbps":171,"fp16_tflops":172,"interconnect_score":155,"parallel_efficiency":173,"score":174,"tier":175,"perf_metric":158,"perf_label":176,"perf_value":177,"workload":10,"cost_metric":178},"RTX 5070",28,48,0.0101,29.49,[149],672,61,0.6,0.5635,"cheapest","98.2 tok\u002Fs",98.2,{"effective_value":179,"effective_unit":158,"capacity_factor":163},68.7,{"gpu_model":181,"gpu_model_id":182,"gpu_count":44,"total_vram_gb":183,"hour_price":184,"monthly_cost":185,"providers":186,"bandwidth_gbps":187,"fp16_tflops":188,"interconnect_score":155,"parallel_efficiency":189,"score":190,"tier":157,"perf_metric":158,"perf_label":191,"perf_value":192,"workload":10,"cost_metric":193},"H200 SXM",11,1128,3.5,20440,[152],4800,989,0.4,0.2197,"934.9 tok\u002Fs",934.9,{"effective_value":194,"effective_unit":158,"capacity_factor":163},654.4,{"gpu_model":196,"gpu_model_id":197,"gpu_count":40,"total_vram_gb":167,"hour_price":198,"monthly_cost":199,"providers":200,"bandwidth_gbps":201,"fp16_tflops":202,"interconnect_score":155,"parallel_efficiency":40,"score":203,"tier":175,"perf_metric":158,"perf_label":204,"perf_value":205,"workload":10,"cost_metric":206},"RTX A6000",6,0.2874,209.8,[149,152],768,77,0.6437,"46.7 tok\u002Fs",46.7,{"effective_value":207,"effective_unit":158,"capacity_factor":163},32.7,[],false,23,"value",[213,217,226,234,241],{"name":196,"vram_gb":167,"cheapest_hour":198,"cost":214,"provider":149,"estimated_tps":205},{"hourly":198,"daily":215,"monthly":199,"yearly":216},6.9,2517.62,{"name":218,"vram_gb":219,"cheapest_hour":220,"cost":221,"provider":149,"estimated_tps":225},"RTX 5090",32,0.337,{"hourly":220,"daily":222,"monthly":223,"yearly":224},8.09,246.01,2952.12,109.1,{"name":227,"vram_gb":167,"cheapest_hour":228,"cost":229,"provider":149,"estimated_tps":233},"L40S",0.4004,{"hourly":228,"daily":230,"monthly":231,"yearly":232},9.61,292.29,3507.5,52.6,{"name":235,"vram_gb":167,"cheapest_hour":236,"cost":237,"provider":149,"estimated_tps":41},"RTX PRO 6000 WS",0.6689,{"hourly":236,"daily":238,"monthly":239,"yearly":240},16.05,488.3,5859.56,{"name":242,"vram_gb":167,"cheapest_hour":243,"cost":244,"provider":149,"estimated_tps":41},"RTX PRO 6000 S",0.7339,{"hourly":243,"daily":245,"monthly":246,"yearly":247},17.61,535.75,6428.96,{"posts_count":41,"benchmarks_count":41}]