Recently, Baidu Famous Agent 2.0, an enterprise-grade autonomous algorithm optimization agent developed by Baidu Smart Cloud, has once again secured the top position on the authoritative machine learning engineering benchmark, MLE-Bench, achieving a new record for peak performance.
Spearheaded by OpenAI, this benchmark features 75 real-world engineering challenges sourced from Kaggle competitions. It is designed to evaluate the end-to-end practical capabilities of artificial intelligence across the entire machine learning workflow, including model training, data preparation, and experiment execution.
Baidu Famous Agent 2.0 shines in tackling "high-difficulty" tasks, showcasing a notably higher overall win rate and surpassing similar agents that leverage mainstream large models.
It is reported that Famous Agent 2.0 has undergone extensive optimizations in areas such as evolutionary strategies, long-term memory mechanisms, and underlying infrastructure. These enhancements have significantly lowered the barrier to algorithm usage, enabling business professionals without an algorithmic background to swiftly generate enterprise-level solutions.
Currently, Famous has garnered attention from thousands of enterprises across core sectors of the national economy, including retail, finance, manufacturing, energy, and transportation. It has also demonstrated immense application value in scientific research scenarios. Its official version is set to be unveiled at the Baidu AI Developer Conference Create 2026 in May this year.