Benchmarks

← All benchmarks

Vending-Bench Arena: Round 13

Games

The Vending-Bench Arena matchup between Claude Opus 5.5, GPT-6 Sol and Grok 4.7, reporting mean final money balances over four runs. Balances are comparable only within this shared-market matchup; the source does not provide a publication day for the round.

Top models (higher is better)