[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-bge-m3":3},{"id":4,"slug":5,"name":6,"hf_id":6,"organization":7,"architecture":8,"architecture_config":9,"parameters":13,"parameters_label":14,"context_length":15,"license":16,"downloads":17,"likes":18,"status":19,"description":20,"huggingface_url":21,"created_at":22,"seo":23,"category":25,"family":28,"score":31,"variants":32,"quantizations":33,"requirements":60,"gpu_recommendations":125,"fitting_gpus":166,"community":201},44,"bge-m3","BAAI\u002Fbge-m3","BAAI","xlm-roberta",{"layers":10,"head_dim":11,"hidden_size":12},24,64,1024,557991936,"558M",8194,null,34918885,3400,"published","BGE-M3 多语言向量模型，8192 上下文，1024 维。","https:\u002F\u002Fhuggingface.co\u002FBAAI\u002Fbge-m3","2026-08-12T02:22:25+00:00",{"title":6,"description":20,"image":16,"robots":24,"canonical_url":16},"index, follow",{"name":26,"slug":27},"Embedding \u002F RAG","embedding",{"name":29,"slug":30},"BGE","bge",0.0924,[],[34,39,44,49,53,57],{"format":35,"bits":36,"size_factor":37,"quality_loss":38},"AWQ",4,0.263,0.03,{"format":40,"bits":41,"size_factor":42,"quality_loss":43},"FP16",16,1,0,{"format":45,"bits":46,"size_factor":47,"quality_loss":48},"FP8",8,0.5,0.01,{"format":50,"bits":36,"size_factor":51,"quality_loss":52},"GGUF",0.25,0.05,{"format":54,"bits":36,"size_factor":55,"quality_loss":56},"GPTQ",0.275,0.04,{"format":58,"bits":46,"size_factor":47,"quality_loss":59},"INT8",0.02,{"FP16":61,"FP8":82,"INT8":93,"AWQ":97,"GPTQ":108,"GGUF":115},{"minimum":62,"production":71,"recommended":76},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":65,"kv_cache_gb":43,"vram_gb":66,"ram_gb":67,"disk_gb":68,"requirement_source":69,"confidence":70},"minimum",4096,1.06,2.72,32,40.65,"estimated",0.95,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":65,"kv_cache_gb":43,"vram_gb":74,"ram_gb":67,"disk_gb":75,"requirement_source":69,"confidence":70},"production",16384,3.33,105.65,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":65,"kv_cache_gb":43,"vram_gb":80,"ram_gb":67,"disk_gb":81,"requirement_source":69,"confidence":70},"recommended",8192,2,2.96,66.65,{"minimum":83,"production":87,"recommended":90},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":84,"kv_cache_gb":43,"vram_gb":85,"ram_gb":67,"disk_gb":86,"requirement_source":69,"confidence":70},0.53,1.97,39.83,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":84,"kv_cache_gb":43,"vram_gb":88,"ram_gb":67,"disk_gb":89,"requirement_source":69,"confidence":70},2.51,104.83,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":84,"kv_cache_gb":43,"vram_gb":91,"ram_gb":67,"disk_gb":92,"requirement_source":69,"confidence":70},2.17,65.83,{"minimum":94,"production":95,"recommended":96},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":84,"kv_cache_gb":43,"vram_gb":85,"ram_gb":67,"disk_gb":86,"requirement_source":69,"confidence":70},{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":84,"kv_cache_gb":43,"vram_gb":88,"ram_gb":67,"disk_gb":89,"requirement_source":69,"confidence":70},{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":84,"kv_cache_gb":43,"vram_gb":91,"ram_gb":67,"disk_gb":92,"requirement_source":69,"confidence":70},{"minimum":98,"production":102,"recommended":105},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":99,"kv_cache_gb":43,"vram_gb":100,"ram_gb":67,"disk_gb":101,"requirement_source":69,"confidence":70},0.27,1.6,39.43,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":99,"kv_cache_gb":43,"vram_gb":103,"ram_gb":67,"disk_gb":104,"requirement_source":69,"confidence":70},2.11,104.43,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":99,"kv_cache_gb":43,"vram_gb":106,"ram_gb":67,"disk_gb":107,"requirement_source":69,"confidence":70},1.79,65.43,{"minimum":109,"production":112,"recommended":113},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":110,"kv_cache_gb":43,"vram_gb":111,"ram_gb":67,"disk_gb":101,"requirement_source":69,"confidence":70},0.28,1.61,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":110,"kv_cache_gb":43,"vram_gb":103,"ram_gb":67,"disk_gb":104,"requirement_source":69,"confidence":70},{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":110,"kv_cache_gb":43,"vram_gb":114,"ram_gb":67,"disk_gb":107,"requirement_source":69,"confidence":70},1.8,{"minimum":116,"production":119,"recommended":122},{"scenario":63,"context_length":64,"batch_size":42,"weight_gb":110,"kv_cache_gb":43,"vram_gb":117,"ram_gb":67,"disk_gb":118,"requirement_source":69,"confidence":70},1.62,39.44,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":110,"kv_cache_gb":43,"vram_gb":120,"ram_gb":67,"disk_gb":121,"requirement_source":69,"confidence":70},2.12,104.44,{"scenario":77,"context_length":78,"batch_size":79,"weight_gb":110,"kv_cache_gb":43,"vram_gb":123,"ram_gb":67,"disk_gb":124,"requirement_source":69,"confidence":70},1.81,65.44,{"recommended":126,"cheapest":142,"performance":145,"min_complexity":155,"cluster_options":163,"no_match":164,"required_vram_gb":106,"candidate_count":36,"usage":165},{"gpu_model":127,"gpu_model_id":10,"gpu_count":42,"total_vram_gb":46,"hour_price":128,"monthly_cost":129,"providers":130,"bandwidth_gbps":132,"fp16_tflops":10,"interconnect_score":133,"parallel_efficiency":42,"score":134,"tier":135,"perf_metric":136,"perf_label":137,"perf_value":138,"workload":16,"cost_metric":139},"RTX 5060 Ti",0.0719,52.49,[131],"vast",448,40,0.5518,"cheapest","vectors\u002Fs","3319 vec\u002Fs",3319,{"effective_value":140,"effective_unit":136,"capacity_factor":141},2655.2,0.8,{"gpu_model":127,"gpu_model_id":10,"gpu_count":42,"total_vram_gb":46,"hour_price":128,"monthly_cost":129,"providers":143,"bandwidth_gbps":132,"fp16_tflops":10,"interconnect_score":133,"parallel_efficiency":42,"score":134,"tier":135,"perf_metric":136,"perf_label":137,"perf_value":138,"workload":16,"cost_metric":144},[131],{"effective_value":140,"effective_unit":136,"capacity_factor":141},{"gpu_model":127,"gpu_model_id":10,"gpu_count":46,"total_vram_gb":11,"hour_price":128,"monthly_cost":146,"providers":147,"bandwidth_gbps":132,"fp16_tflops":10,"interconnect_score":133,"parallel_efficiency":148,"score":149,"tier":150,"perf_metric":136,"perf_label":151,"perf_value":152,"workload":16,"cost_metric":153},419.9,[131],0.4,0.3615,"balanced","26548 vec\u002Fs",26548,{"effective_value":154,"effective_unit":136,"capacity_factor":141},21238.4,{"gpu_model":127,"gpu_model_id":10,"gpu_count":79,"total_vram_gb":41,"hour_price":128,"monthly_cost":156,"providers":157,"bandwidth_gbps":132,"fp16_tflops":10,"interconnect_score":133,"parallel_efficiency":141,"score":158,"tier":135,"perf_metric":136,"perf_label":159,"perf_value":160,"workload":16,"cost_metric":161},104.97,[131],0.4504,"6637 vec\u002Fs",6637,{"effective_value":162,"effective_unit":136,"capacity_factor":141},5309.6,[],false,"value",[167,174,182,189,196],{"name":168,"vram_gb":41,"cheapest_hour":169,"cost":170,"provider":131,"estimated_tps":43},"Tesla V100",0.0289,{"hourly":169,"daily":171,"monthly":172,"yearly":173},0.69,21.1,253.16,{"name":175,"vram_gb":10,"cheapest_hour":176,"cost":177,"provider":131,"estimated_tps":181},"RTX 3090",0.0622,{"hourly":176,"daily":178,"monthly":179,"yearly":180},1.49,45.41,544.87,3466.7,{"name":183,"vram_gb":41,"cheapest_hour":184,"cost":185,"provider":131,"estimated_tps":43},"RTX 4070S Ti",0.0678,{"hourly":184,"daily":186,"monthly":187,"yearly":188},1.63,49.49,593.93,{"name":190,"vram_gb":41,"cheapest_hour":191,"cost":192,"provider":131,"estimated_tps":43},"RTX 4080S",0.0685,{"hourly":191,"daily":193,"monthly":194,"yearly":195},1.64,50.01,600.06,{"name":127,"vram_gb":46,"cheapest_hour":128,"cost":197,"provider":131,"estimated_tps":200},{"hourly":128,"daily":198,"monthly":129,"yearly":199},1.73,629.84,1659.3,{"posts_count":43,"benchmarks_count":43}]