Behind the Benchmarks: Building Agent Leaderboard 2.0

​Join the Galileo team for an exclusive livestream walkthrough of the latest updates to the Galileo Agent Leaderboard—the benchmark for evaluating the real-world performance of AI agents.

​We’ll unpack what’s changed (and why it matters), spotlight emerging leaders across evaluation categories like planning, tool use, and multi-turn reasoning, and reveal key insights.

​This session is perfect for:

• ​AI engineers working with agentic systems

• ​Builders evaluating complex multi-step agents

• ​Researchers curious about the edge of capability evaluations

• ​Teams interested in how self-reflection and domain adherence are scored at scale

​🔍 What you'll learn:

• ​How Galileo measures agent reliability in the wild

• ​Behind-the-scenes updates

• ​Surprising model performance shifts and new failure modes

• ​What's next for the leaderboard

• ​How to make the most of this resource.

​🎤 Feat: Creator of the Leaderboard, Pratik Bhavasar + Sr. Developer Experience Engineer, Erin Mikail Staples.

​💬 Live Q&A and leaderboard AMA at the end.