GPT-5.6 Luna Outperforms Sol on Vercel Leaderboard After 80% Price Cut

cramforce · x · 2026-07-31

Citing Vercel AI Gateway's DeepsecBench (application security vulnerability detection) leaderboard, developer @cramforce pointed out that following an 80% price reduction, GPT-5.6 Luna running on xhigh now outperforms GPT-5.6 Sol on medium.

According to the benchmark data, Luna (xhigh) ranks 4th with a score of 26.79 at a cost of $24.59, while Sol (medium) scores 25.10 at a lower cost ($17.95). This highlights Luna as a highly cost-effective option after the significant price cut.

Related event: GPT-5.6 Luna Offers Unbeatable Cost-Performance(3 posts)→

Original post →

More from Models

Models channel →