Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Agent Skill Performance on SpreadsheetBench
Loading...
69.83
Vrf Score
Trace2Skill (Human-Written + Combined)
18.5268
31.8459
45.165
58.4841
Mar 26, 2026
Vrf Score
Soft Score
Hard Score
Updated 22d ago
Evaluation Results
Method
Method
Links
Vrf Score
Soft Score
Hard Score
Trace2Skill (Human-Written + Combined)
Skill User=122B, Deepe...
2026.03
69.83
47.17
29.53
ReasoningBank
Skill User=122B, Deepe...
2026.03
56
40.1
21.3
Trace2Skill (Human-Written + Combined)
Skill User=35B, Deepen...
2026.03
29.67
18.8
5.73
ReasoningBank
Skill User=35B, Deepen...
2026.03
20.5
17.3
4.97
Feedback
Search any
task
Search any
task