Versions

1 version · Metric: Average test pass rate

Versions of NL2Repo Bench
VersionReleasedModelsResults
NL2Repo Bench OriginalLatestDecember 14, 202555

Recently reported results: NL2Repo Bench Original

Latest version: NL2Repo Bench Original

Recently reported results for NL2Repo Bench Original
ModelScoreSource
GLM-5.3-Flash (max)56.3%autoclaw.z.ai for GLM-5.3-Flash (max) on NL2Repo Bench Original (opens in a new tab)Average test pass rateReported Launch postAutoClaw Team (Z.ai)Self-reportedZ.aiRecord for GLM-5.3-Flash on NL2Repo Bench Original
GLM-5.3 (max)58.0%z.ai for GLM-5.3 (max) on NL2Repo Bench Original (opens in a new tab)Average test pass rateReported Launch postZ.aiSelf-reportedRecord for GLM-5.3 on NL2Repo Bench Original
GLM-5.248.9%z.ai for GLM-5.2 on NL2Repo Bench Original (opens in a new tab)Average test pass rateReported Launch postZ.aiSelf-reportedRecord for GLM-5.2 on NL2Repo Bench Original
Qwen3.6-35B-A3B29.4%huggingface.co for Qwen3.6-35B-A3B on NL2Repo Bench Original (opens in a new tab)Average test pass rateReported Model cardQwenSelf-reportedQwen TeamRecord for Qwen3.6-35B-A3B on NL2Repo Bench Original
GLM-5.142.7%z.ai for GLM-5.1 on NL2Repo Bench Original (opens in a new tab)Average test pass rateReported Launch postZ.aiSelf-reportedRecord for GLM-5.1 on NL2Repo Bench Original