PostHog MCP Leaderboard (posthog.com)

Last week, PostHog released its latest MCP Leaderboard, showcasing the weekly usage share of various AI models in tool calls. Highlighting significant growth since June 22, the data reveals that the top performers include models like Claude Opus 5, GPT-6, and Claude Sonnet 5. Each model’s usage is tracked meticulously, indicating not only which models are most utilized but also providing insight into their performance related to error rates and tool schema adherence. Claude models, for example, made up a considerable share of tool calls, reflecting their growing acceptance in the AI tool ecosystem. This leaderboard is significant for the AI/ML community as it provides a transparent view of the competitive landscape in model usage and performance metrics. By analyzing usage patterns and error rates, developers and researchers can identify strengths and weaknesses in different AI models, informing future developments and optimizations. The emphasis on compliance with the MCP protocol further highlights the importance of model accuracy and responsiveness, as well as the intricate relationship between AI models and the environments they operate in.