List your company

Software Engineering on SWE-bench Verified (resolution rate)

83.8Resolution RateMemCoder
-3.2417619.3556241.95364.55038Dec 16, 2025Jan 12, 2026Feb 8, 2026Mar 8, 2026Apr 4, 2026May 1, 2026May 29, 2026

Evaluation results

MethodLinks
MemCoderSetting=pass@2, Backbone=GPT-5.22026.0283.8—419
Claude Opus 4.6Scaffold=Claude Code2026.0380.8——
KAT-Coder-V2Scaffold=Claude Code2026.0379.6——
MemCoderSetting=pass@1, Backbone=GPT-5.22026.0278.8—394
MemCoderSetting=pass@1, Backbone=DeepSeek-V3.22026.0277.8—389
OpenHandsSetting=pass@3, Backbone=Claude Opus 4.52026.0277.6—388
Claude Opus 4.6Scaffold=OpenClaw2026.0375.7——
Claude Opus 4.6Scaffold=OpenCode2026.0375——
KAT-Coder-V2Scaffold=OpenCode2026.0374.8——
OpenHandsSetting=pass@3, Backbone=Claude Sonnet 4.52026.0274.6—373
OpenHandsSetting=pass@3, Backbone=GPT-5.22026.0274.4—372
KAT-Coder-V2Scaffold=OpenClaw2026.0372.8——
OpenHandsSetting=pass@3, Backbone=Gemini 3 pro2026.0270.4—352
Full-ContextAgent Backbone=GLM-4.7, Latency (s)=5869.42, Speedup=1.00×2026.0569——
COMEM (GRPO)Agent Backbone=GLM-4.7, Latency (s)=2821.38, Speedup=2.08×2026.0562.7——
Full-ContextAgent Backbone=Qwen3-Coder-Max, Latency (s)=5129.90, Speedup=1.00×2026.0557.2——
MemAgent (7B)Agent Backbone=GLM-4.7, Latency (s)=3253.83, Speedup=1.80×2026.0554.8——
COMEM (GRPO)Agent Backbone=Qwen3-Coder-Max, Latency (s)=3594.51, Speedup=1.43×2026.0551——
MemAgent (7B)Agent Backbone=Qwen3-Coder-Max, Latency (s)=3672.48, Speedup=1.40×2026.0543.2——
Proxy-filtered (code survival = 1)Selection Strategy=Proxy-filtered (code survival = 1)2026.030.5043.8—
Critic-selectedSelection Strategy=Critic-selected2026.030.4781.2—
Base Model (no SFT)Selection Strategy=Base Model (no SFT)2026.030.466——
RandomSelection Strategy=Random2026.030.462-0.4—
SWE-smith-LM-32BScaffold=SWE-Agent, Train size=5,016, Evaluation=Reported2025.120.402——
OEC-SWE-32BScaffold=SWE-Agent, Train size=4,961, Evaluation=Own evaluation2025.120.4——
Skywork-SWE-32BScaffold=OpenHands, Train size=8,4472025.120.38——
SWE-smith-LM-32BScaffold=SWE-Agent, Train size=5,016, Evaluation=Own evaluation2025.120.364——
OEC-SWE-7BScaffold=SWE-Agent, Train size=8,943, Evaluation=Own evaluation2025.120.208——
SWE-Gym-32BScaffold=OpenHands, Train size=4912025.120.206——
SWE-smith-LM-7BScaffold=SWE-Agent, Train size=5,0162025.120.152——
SkyRL-Agent-7B-v0Scaffold=OpenHands, Train size=3,680+2025.120.146——
SWE-Gym-7BScaffold=OpenHands, Train size=4912025.120.106——