Claude 3 family matches GPT-4

Anthropic's Claude 3 Opus tops GPT-4 on several reasoning benchmarks including GPQA.

What happened and why it matters

Claude 3 Opus ended GPT-4's year-long sole frontier status, posting ~50% on the graduate-level GPQA Diamond and strong long-context recall. Multiple labs reaching the frontier signalled the capability was reproducible, not a one-off.

Sources

• Introducing the next generation of Claude ↗ Anthropic · 4 Mar 2024