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

Deep Research on HealthBench

59.5Score

GPT-5 + Search

-15.93.67523.2542.825Nov 24, 2025Dec 27, 2025Jan 29, 2026Mar 3, 2026Apr 5, 2026May 8, 2026Jun 10, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.11
59.5
2026.06
59.5
2026.06
54.4
2025.11
53.8
2026.06
53.8
2025.11
52.8
2026.06
52.1
2026.06
51
2026.06
50.8
2026.06
50.2
2026.06
50.2
2026.06
49.6
2025.11
46.2
2026.06
46.2
2025.11
39.4
2026.06
39.4
2025.11
38.1
2026.06
38.1
2025.11
38
2026.06
38
2025.11
36.5
2026.06
36.5
2025.11
33.7
2026.06
33.7
2025.11
32
2026.06
32
2025.11
31.1
2025.11
24.5
2026.06
21.3
2025.11
16.5
2025.11
11.1
2026.06
11.1
2025.11
5.9
2026.06
5.9
2025.11
-0.1
2026.06
-0.1
2025.11
-13
2026.06
-13