🛠 Claude Code Usage Leaderboard on claudeusage.com
The service published data from 130 public profiles: total user costs amounted to nearly $57,400, with a volume of 3.8 billion tokens and 10.7 thousand hours of active work. The top models are dominated by Claude Sonnet (4.5, 4.6) and Claude Opus (4.7, 4.8).
🌍 It demonstrates the real scale and intensity of using agentic development tools (Claude Code) in workflows, as well as the popularity of Anthropic models.
👤 It allows for an assessment of the economics of using advanced AI agents and shows which models are most in demand in intensive development scenarios.
Source 1: https://www.claudeusage.com/leaderboard