Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Tool Use Task Completion on MCP Tool-Use Tasks (evaluation)
Loading...
89.9
Task Success Rate (TSR)
TRUSTDESC
83.868
85.434
87
88.566
Apr 8, 2026
Task Success Rate (TSR)
Completed Tasks Count
Updated 9d ago
Evaluation Results
Method
Method
Links
Task Success Rate (TSR)
Completed Tasks Count
TRUSTDESC
Model for Gen=Claude-4...
2026.04
89.9
187
TRUSTDESC
Model for Gen=Gemini-3...
2026.04
88.5
184
TRUSTDESC (Average)
Model for Gen=Average,...
2026.04
87.7
182.5
TRUSTDESC
Model for Gen=Claude-4...
2026.04
87.5
182
TRUSTDESC
Model for Gen=GPT-5.2,...
2026.04
87
181
TRUSTDESC
Model for Gen=Gemini-3...
2026.04
86.5
180
TRUSTDESC
Model for Gen=gpt-oss-...
2026.04
86.5
180
TRUSTDESC (Average)
Model for Gen=Average,...
2026.04
86.3
179.5
TRUSTDESC
Model for Gen=GPT-5.2,...
2026.04
86.1
179
Original Descriptions
Description Type=original
2026.04
84.1
175
TRUSTDESC
Model for Gen=gpt-oss-...
2026.04
84.1
175
Feedback
Search any
task
Search any
task