ResearchBenchmarksCode on APT-BenchFollow41.9AccuracyQwen322.86827.80932.7537.691Dec 31, 2025Evaluation ResultsMethodMethodLinksAccuracyQwen3Parameters=4B, Variant...Parameters=4B, Variant=Base2025.1241.9Youtu-LLMParameters=2B, Variant...Parameters=2B, Variant=Base2025.1237.9Gemma3Parameters=4B, Variant...Parameters=4B, Variant=Base2025.1232.8Qwen3Parameters=1.7B, Varia...Parameters=1.7B, Variant=Base2025.1225.1SmolLM3Parameters=3B, Variant...Parameters=3B, Variant=Base2025.1224.3Llama3.1Parameters=8B, Variant...Parameters=8B, Variant=Base2025.1223.6