[{"rank":1,"reference":false,"run_id":"a4cc8ae1a524","task":"clinicalqa@0","model":{"slug":"kgpu/glm-5.3-ax-753b-a40b","name":"GLM-5.3 753B-A40B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.9090909090909091,"ci_low":0.8918421052631579,"ci_high":0.9263157894736842},"n":1045,"parse_rate":0.9980861244019139,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/a4cc8ae1a524"},{"rank":2,"reference":false,"run_id":"f7b3c2b141f5","task":"clinicalqa@0","model":{"slug":"openrouter/google/gemini-3.8-flash","name":"Gemini 3.8 Flash","type":"llm","provider":"openrouter"},"tier":"baseline","metric":{"name":"accuracy","value":0.908133971291866,"ci_low":0.8908851674641148,"ci_high":0.9253588516746412},"n":1045,"parse_rate":0.999043062200957,"cost_usd":2.3561491,"eval_date":"2026-10-29","url":"https://medbench.kr/r/f7b3c2b141f5"},{"rank":3,"reference":false,"run_id":"a988457822dd","task":"clinicalqa@0","model":{"slug":"kgpu/qwen3.8-max-ax-2.4t-a95b","name":"Qwen3.8 Max 2.4T-A95B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.9062200956937799,"ci_low":0.8889952153110048,"ci_high":0.9234449760765551},"n":1045,"parse_rate":0.9942583732057416,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/a988457822dd"},{"rank":4,"reference":false,"run_id":"2224a6c5f8d6","task":"clinicalqa@0","model":{"slug":"openrouter/deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash (2026-10-17)","type":"llm","provider":"openrouter"},"tier":"baseline","metric":{"name":"accuracy","value":0.9043062200956937,"ci_low":0.8870813397129187,"ci_high":0.9224880382775119},"n":1045,"parse_rate":0.9980861244019139,"cost_usd":0.47307404999999997,"eval_date":"2026-10-29","url":"https://medbench.kr/r/2224a6c5f8d6"},{"rank":5,"reference":false,"run_id":"d982ddb4d0cc","task":"clinicalqa@0","model":{"slug":"kgpu/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.9033492822966507,"ci_low":0.8851674641148325,"ci_high":0.9215311004784689},"n":1045,"parse_rate":0.9980861244019139,"cost_usd":2.28150645,"eval_date":"2026-10-29","url":"https://medbench.kr/r/d982ddb4d0cc"},{"rank":6,"reference":false,"run_id":"70314bc6d7e9","task":"clinicalqa@0","model":{"slug":"kgpu/deepseek-v4.1-flash-ax-552b-a16b","name":"DeepSeek V4.1 Flash 552B-A16B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.9004784688995215,"ci_low":0.8832535885167464,"ci_high":0.9177033492822967},"n":1045,"parse_rate":0.9971291866028709,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/70314bc6d7e9"},{"rank":7,"reference":false,"run_id":"d15ecb67085b","task":"clinicalqa@0","model":{"slug":"openrouter/openai/gpt-6-luna","name":"GPT-6 Luna","type":"llm","provider":"openrouter"},"tier":"baseline","metric":{"name":"accuracy","value":0.9004784688995215,"ci_low":0.8822966507177034,"ci_high":0.9186602870813397},"n":1045,"parse_rate":1.0,"cost_usd":0.11027780000000001,"eval_date":"2026-10-29","url":"https://medbench.kr/r/d15ecb67085b"},{"rank":8,"reference":false,"run_id":"f494ae371950","task":"clinicalqa@0","model":{"slug":"kgpu/deepseek-v4-pro-ax-1.6t-a49b","name":"DeepSeek V4 Pro 1.6T-A49B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.8937799043062201,"ci_low":0.8755980861244019,"ci_high":0.9110047846889953},"n":1045,"parse_rate":1.0,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/f494ae371950"},{"rank":9,"reference":false,"run_id":"f1ff027148b1","task":"clinicalqa@0","model":{"slug":"kgpu/qwen3.8-flash-next-180b-a6b","name":"Qwen3.8 Flash Next 180B-A6B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.8937799043062201,"ci_low":0.876555023923445,"ci_high":0.9129186602870814},"n":1045,"parse_rate":0.9799043062200957,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/f1ff027148b1"},{"rank":10,"reference":false,"run_id":"bc2d61a50470","task":"clinicalqa@0","model":{"slug":"openrouter/anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","type":"llm","provider":"openrouter"},"tier":"baseline","metric":{"name":"accuracy","value":0.8851674641148325,"ci_low":0.8660287081339713,"ci_high":0.9033492822966507},"n":1045,"parse_rate":0.9837320574162679,"cost_usd":0.4408795,"eval_date":"2026-10-29","url":"https://medbench.kr/r/bc2d61a50470"},{"rank":11,"reference":false,"run_id":"33d8b39d1c90","task":"clinicalqa@0","model":{"slug":"kgpu/gpt-oss-120b-a5.1b","name":"gpt-oss 120B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.8822966507177034,"ci_low":0.8631578947368421,"ci_high":0.9014354066985646},"n":1045,"parse_rate":1.0,"cost_usd":0.0720534069,"eval_date":"2026-10-29","url":"https://medbench.kr/r/33d8b39d1c90"},{"rank":12,"reference":false,"run_id":"d3265cd5bd82","task":"clinicalqa@0","model":{"slug":"kgpu/gemma-4-26b-a4b","name":"Gemma 4 26B-A4B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.8736842105263158,"ci_low":0.8535885167464115,"ci_high":0.8937799043062201},"n":1045,"parse_rate":0.999043062200957,"cost_usd":0.20353827000000002,"eval_date":"2026-10-29","url":"https://medbench.kr/r/d3265cd5bd82"},{"rank":13,"reference":false,"run_id":"75310a96c0a0","task":"clinicalqa@0","model":{"slug":"kgpu/gpt-oss-20b-a3.6b","name":"gpt-oss 20B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.8516746411483254,"ci_low":0.830622009569378,"ci_high":0.8736842105263158},"n":1045,"parse_rate":0.9952153110047847,"cost_usd":0.09091615540000002,"eval_date":"2026-10-29","url":"https://medbench.kr/r/75310a96c0a0"},{"rank":14,"reference":false,"run_id":"7993fc36d3a5","task":"clinicalqa@0","model":{"slug":"kgpu/medgemma-1.5-4b","name":"MedGemma 1.5 4B","type":"llm","provider":"kgpu"},"tier":"baseline","metric":{"name":"accuracy","value":0.7645933014354067,"ci_low":0.738755980861244,"ci_high":0.7894736842105263},"n":1045,"parse_rate":0.9952153110047847,"cost_usd":0.12438564,"eval_date":"2026-10-29","url":"https://medbench.kr/r/7993fc36d3a5"},{"rank":null,"reference":true,"run_id":"a9dc010ea139","task":"clinicalqa@0","model":{"slug":"medbench/majority-b","name":"Majority class (always B)","type":"baseline","provider":"medbench"},"tier":"baseline","metric":{"name":"accuracy","value":0.34258373205741627,"ci_low":0.3138755980861244,"ci_high":0.3703349282296651},"n":1045,"parse_rate":1.0,"cost_usd":0.0,"eval_date":"2026-10-29","url":"https://medbench.kr/r/a9dc010ea139"}]