Introducing the SWE-Lancer benchmark

A new benchmark called SWE-Lancer has been introduced to evaluate whether frontier large language models can earn $1 million through real-world freelance software engineering work. This benchmark tests AI capabilities in practical, revenue-generating programming tasks rather than traditional academic assessments.