Share your thoughts, 1 month free Claude Pro on usSee more

Medical Calculation on MedCalc-Bench

81.4Accuracy

GPT-5

Updated 22d ago

Evaluation Results

Method
GPT-5 2026.05	81.4	-	-	-	-	-	-
Claude Sonnet 4.5 2026.07	76	-	-	-	-	-	-
DeepSeek-R1 2026.07	72.7	-	-	-	-	-	-
GPT-4.1 2026.05	72	-	-	-	-	-	-
GPT-4.1-mini 2026.05	67.5	-	-	-	-	-	-
FaithMed (Qwen3 4B) 2026.07	56.5	-	-	-	-	-	-
MedGemma-27B 2026.05	51.9	-	-	-	-	-	-
MedGemma 27B 2026.07	49	-	-	-	-	-	-
CLR-voyance-8B 2026.05	46	-	-	-	-	-	-
Qwen3 8B 2026.07	44.9	-	-	-	-	-	-
Qwen3-8B base 2026.05	41.5	-	-	-	-	-	-
Agentic Search + GRPO (Qwen3 4B) 2026.07	39	-	-	-	-	-	-
FaithMed (Qwen3 1.7B) 2026.07	33.2	-	-	-	-	-	-
DS-R1-0528 2026.05	31.6	-	-	-	-	-	-
Agentic Search (Qwen3 4B) 2026.07	29.6	-	-	-	-	-	-
Agentic Search + GRPO (Qwen3 1.7B) 2026.07	27.1	-	-	-	-	-	-
HuatuoGPT-o1 2026.05	27	-	-	-	-	-	-
Gemma 3 12B 2026.07	27	-	-	-	-	-	-
Llama 3.1 8B Instruct 2026.07	21.4	-	-	-	-	-	-
MG-B 2026.05	21	-	-	-	-	-	-
MG-B+A 2026.05	18.5	-	-	-	-	-	-
Base (Qwen3 1.7B) 2026.07	16.9	-	-	-	-	-	-
MG-Δ+A 2026.05	15.5	-	-	-	-	-	-
Agentic Search (Qwen3 1.7B) 2026.07	15.3	-	-	-	-	-	-
CLR-voyance-4B 2026.05	15	-	-	-	-	-	-
Base (Qwen3 4B) 2026.07	13.7	-	-	-	-	-	-
MG-4B base 2026.05	10.7	-	-	-	-	-	-
MedRAG (Qwen3 4B) 2026.07	10.3	-	-	-	-	-	-
DS-R1-Distill 2026.05	8.6	-	-	-	-	-	-
MedRAG (Qwen3 1.7B) 2026.07	8.4	-	-	-	-	-	-
Gemma-3-4B 2026.05	7.3	-	-	-	-	-	-
Empty CoT 2026.01	-	0.1	0.18	0.45	0.36	0.38	0.29
Default (Greedy) CoT 2026.01	-	0.14	0.24	0.43	0.38	0.41	0.32
Default (Sampled) CoT 2026.01	-	0.16	0.29	0.45	0.38	0.35	0.33
Trans. CoT (NRR) 2026.01	-	0.14	0.15	0.24	0.3	0.34	0.23
Trans. CoT (OpenT) 2026.01	-	0.21	0.24	0.26	0.26	0.25	0.24
Trans. CoT (OSS) 2026.01	-	0.26	0.44	0.43	0.4	0.44	0.39
Trans. CoT (QwQ) 2026.01	-	0.34	0.37	0.39	0.38	0.37	0.37
Trans. CoT (DAPO) 2026.01	-	0.31	0.4	0.39	0.4	0.41	0.38
Ens. (QwQ+DAPO/OSS) 2026.01	-	0.39	0.37	0.41	0.41	0.4	0.4
Ens. (QwQ+OSS/DAPO) 2026.01	-	0.28	0.37	0.45	0.42	0.43	0.38
Ens. (QwQ+OpenT/OSS) 2026.01	-	0.34	0.41	0.42	0.38	0.4	0.39