AI Benchy is a platform for comparing AI models through an interactive benchmark leaderboard. It evaluates hundreds of models and helps users assess their scores, reliability across test runs, response times, costs, and overall value. The leaderboard includes models from providers such as Google and OpenAI, alongside offerings from Anthropic, DeepSeek, Qwen, and others.

Users can filter results by score, cost, response time, correct tests, model availability, and architecture, then compare selected models side by side. AI Benchy also provides interactive charts to explore trade-offs and decision-focused rankings covering areas such as routing, policy, evidence, and resistance to prompt injection. Model pages and comparison links make it easier to examine individual entries and review head-to-head matchups.

The site includes showcases and an AI World Cup, as well as language and theme options. By bringing benchmark results and practical comparison tools together, AI Benchy helps users make more informed choices among AI models based on performance, speed, reliability, and price