The AI arena is free today

Open Superagent

Methodology

Everything LLM Stats publishes — leaderboards, benchmark scores, pricing, speed, and composite indexes — is derived from public sources and independently sampled measurements. This page summarizes the principles behind each metric.

Sources we rely on

  • Provider-published pricing, checked against observed billing data.
  • Verified benchmark releases (lab-published or independently replicated). Unverified scores are flagged in the UI and excluded from composite indexes.
  • Live performance measurements collected consistently across supported providers.
  • Community arenas with quality controls designed to protect the integrity of preference data.

For additional methodology information, please contact us.