Using Arena's data, we take a deep dive into model performance across the leading labs. We'll explore how the race has evolved, who's winning overall, which models offer the best performance at a given price point, and who's leading in open source.

Peter is an AI Capability Lead at Arena, focused on testing and benchmarking model performance. He was previously Head of AI at UK e-commerce company Moonpig, where he led AI adoption across the business, following earlier AI and data science roles in banking and consulting.