Baidu ERNIE Agent tops PinchBench v2 benchmark
Baidu
Baidu ERNIE Agent achieved first place in the well-known large language model benchmark PinchBench v2, demonstrating superior performance in agent tasks.
Baidu's ERNIE Agent model has topped the leaderboard of PinchBench v2, a prominent benchmark for evaluating large language models. The benchmark assesses agent capabilities, including planning, tool use, and multi-step reasoning. ERNIE Agent outperformed other models, showcasing Baidu's strength in AI agent development.
Source: Baidu ERNIE (GNews) —
original