Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
URL Health and Self-Correction on ExpertQA
Loading...
7,985
Total URLs
Claude
4,051.72
5,072.86
6,094
7,115.14
Apr 3, 2026
Total URLs
Avg URLs per Question
Live URL Rate
Dead URL Rate
Likely Hallucination Rate
Unknown URL Status Rate
Avg Live URLs per Question
Mean Correction Rounds
Median Correction Rounds
Max Correction Rounds
Updated 1mo ago
Evaluation Results
Method
Method
Links
Total URLs
Avg URLs per Question
Live URL Rate
Dead URL Rate
Likely Hallucination Rate
Unknown URL Status Rate
Avg Live URLs per Question
Mean Correction Rounds
Median Correction Rounds
Max Correction Rounds
Claude
Agentic tool usage=url...
2026.04
7,985
18.4
79.3
10
0.4
20.2
14.6
4.06
4
8
GPT-5.1
Agentic tool usage=url...
2026.04
4,829
11.1
78
60
1.8
19.7
8.7
2.52
2
6
Gemini
Agentic tool usage=url...
2026.04
4,203
9.7
88.9
20
0.5
10.3
8.6
1.56
2
2
Feedback
Search any
task
Search any
task
