Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Software Engineering Performance on SWE-bench Lite (Accuracy)
Loading...
16.1
Accuracy
TOOLSELF
9.86
11.48
13.1
14.72
Feb 8, 2026
Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy
TOOLSELF
Base Model=Qwen3-14B
2026.02
16.1
SWE-Search
Base Model=Qwen3-14B
2026.02
14.6
SWE Agent
Base Model=Qwen3-14B
2026.02
14.2
Vanilla Agent
Base Model=Qwen3-14B
2026.02
13.3
TOOLSELF
Base Model=Qwen3-8B
2026.02
12.4
SWE-Search
Base Model=Qwen3-8B
2026.02
11.6
SWE Agent
Base Model=Qwen3-8B
2026.02
10.9
Vanilla Agent
Base Model=Qwen3-8B
2026.02
10.1
Feedback
Search any
task
Search any
task