New: Claude Fable 5.2 eval report is here.Read the report ↗

Best models for running a business

Vending-Bench 2

Measures long-term business management through a simulated vending business; higher final bank balance is better.

Updated: Data as of:
Limited reportingSaturation: not applicable
Official benchmark
Vending-Bench 2 results. Each row identifies its benchmark version and evaluation track. Scores from different versions are not directly comparable. Best effort from the latest stored report date per model and version.
Model / creatorVersionReported score
Gemini 4 HydrogenGoogle · Proprietary2 / Google Hydrogen launch comparison14,549 USD
Claude Sonnet 5.5Anthropic · Proprietary2 / Google Hydrogen launch comparison11,165 USD
GPT-6.1 TerraOpenAI · Proprietary2 / Google Hydrogen launch comparison9,107 USD
View all 3 published reports
2 / Google Hydrogen launch comparison

effort: highest available or best reported · Google quotes Andon Labs public leaderboard final bank balance; values are dollars, not percentages. · Tools enabled

Google DeepMind ↗
2 / Google Hydrogen launch comparison

effort: highest available or best reported · Google quotes Andon Labs public leaderboard final bank balance; values are dollars, not percentages. · Tools enabled

Google DeepMind ↗
GPT-6.1 Terra9,107 USD
2 / Google Hydrogen launch comparison

effort: highest available or best reported · Google quotes Andon Labs public leaderboard final bank balance; values are dollars, not percentages. · Tools enabled

Google DeepMind ↗