Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Agentic (multi-turn) evaluation on NitiBench
Loading...
78.02
Accuracy
InK-GRPO
36.472
47.2585
58.045
68.8315
Jan 26, 2026
Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy
InK-GRPO
agentic=true
2026.01
78.02
GPT-5
agent=true
2026.01
75.34
GRPO
agentic=true
2026.01
73.73
Qwen3-4B-Instruct-2507
agent=true
2026.01
46.11
GPT-5
search=true
2026.01
38.07
Feedback
Search any
task
Search any
task