SWE-BENCH
SWE-Bench is a leader board that evaluates AI models by testing their ability to solve real-world coding tasks from GitHub issues. This is a collaboration between Princeton and Standford Universities.
Short Name
SWE-BENCH
Category
Web
SubCategory
miscellaneous
Risk
2
Release Date
2025-09-09
Supported Platforms
Available to all SRX running 12.3X48+
Available to all MX running 20.2R1+
Available to all vSRX running 20.3R1+
Port
UDP/53,TCP/25,TCP/22,TCP/8080,TCP/1080,TCP/21,TCP/465,UDP/500,UDP/135,UDP/554,TCP/1434,UDP/1900,TCP/102,TCP/135,UDP/5246-5247,TCP/119,TCP/53,TCP/139,UDP/3544,TCP/5445,TCP/5190,TCP/3386,TCP/434,TCP/3128,TCP/80,TCP/587,UDP/80,TCP/995,TCP/5060,UDP/4500,TCP/1433,UDP/443,TCP/20,TCP/993,TCP/543,TCP/445,TCP/8000,UDP/5060,TCP/443,UDP/3386,TCP/554,UDP/88
Application Group
junos-approot:applications:ai