Eseandre Mordi
⬤ Grok 4 continues crushing it on the Vending-Bench leaderboard, holding the top spot with performance numbers that blow away the competition. Even with tons of new models hitting the market, nothing's managed to touch Grok 4's results. The latest data shows just how much it's pulled ahead across every key metric.
⬤ The numbers tell the story: Grok 4 posted a mean net worth of $4,694.15 and a minimum of $3,333.28—way ahead of everyone else. GPT-5 sits in second with $3,578.90, while Claude Opus 4 trails at $2,077.41. Grok 4 also dominated in total units sold, hitting 4,569 units versus GPT-5's 2,471 and Claude Opus 4's 1,412. That's nearly double the closest competitor.
⬤ Looking at sustained performance, Grok 4 kept selling for 324 days straight—99.5% of its run duration. Other models showed mixed results: Claude Sonnet 4.5 had decent minimum sales but couldn't match profitability, while Gemini 2.5 Pro, o3, and others lagged in both net worth and sales duration. Nothing in the rankings came close to Grok 4's combination of sales power, longevity, and profit generation.
⬤ What makes this interesting is how clearly benchmark performance separates the winners from the rest in today's crowded AI landscape. Grok 4's continued dominance shows real market traction and highlights the growing gap between top-tier models and everything else. As more systems get evaluated on platforms like Vending-Bench, competitive positioning increasingly comes down to actual measurable results rather than just hype.
Eseandre Mordi
