Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Software Engineering Task Resolution on SWE-bench In-Distribution
Loading...
52.34
Resolved Rate
SkillFlow
14.9728
24.6739
34.375
44.0761
May 13, 2026
Resolved Rate
Updated 19d ago
Evaluation Results
Method
Method
Links
Resolved Rate
SkillFlow
Method Category=Ours
2026.05
52.34
FlowSteer
Method Category=Agent+RL
2026.05
43.75
AgentFlow
Method Category=Agent+RL
2026.05
41.41
SkillRL
Method Category=Agent+RL
2026.05
39.06
v4-flash
Method Category=Baseline
2026.05
38.28
Qwen3.5-AFlow
Method Category=AFlow,...
2026.05
30.47
Qwen3.5-GRPO
Method Category=GRPO,...
2026.05
18.75
Qwen3.5-9B
Method Category=Baseline
2026.05
17.19
Qwen3.5-SFT
Method Category=SFT, B...
2026.05
16.41
Feedback
Search any
task
Search any
task