Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Workflow Reconstruction on METAGPT (test)
Loading...
55.7
SFE
AgentXRay (All Tools)
26.06
33.755
41.45
49.145
Feb 5, 2026
SFE
Updated 4d ago
Evaluation Results
Method
Method
Links
SFE
AgentXRay (All Tools)
Config=Full primitive...
2026.02
55.7
AgentXRay (Selected)
Config=Selected primit...
2026.02
47
AgentXRay w/o Pruning
Ablation=Without Pruning
2026.02
33.4
ReAct (Claude Opus 4.5)
Protocol=ReAct-style t...
2026.02
33.1
Claude Opus 4.5
Protocol=Multi-turn se...
2026.02
32.2
AgentXRay w/o Tools
Ablation=Without Tools
2026.02
30.1
AFlow
Protocol=MCTS-based wo...
2026.02
28
SFT
Protocol=Direct input-...
2026.02
27.2
Feedback
Search any
task
Search any
task