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

Seal-Tools

Benchmarks

Task NameDataset NameSOTA ResultTrend
Tool CallingSeal-Tools Single-Tool
Name Match Score98.14
30
Tool CallingSeal-Tools Single-Tool v1 (test)
F1 Name98.14
12
Function CallingSeal-Tools
F1 Score94.94
11
Tool SelectionSeal-Tools (test)
Top-1 Acc99.9
2
Agent-team recommendationSeal-Tools (test)
Retrieval Success (Top-10)99.6
2
Showing 5 of 5 rows