# Vending-Bench 2

An AI runs a simulated vending-machine business for a long time; tests staying sensible over many days.

Category: agents. Type: leaderboard. Status: active.
Metric: Average final money balance ($) (higher is better).
Source: Andon Labs — https://andonlabs.com/evals/vending-bench-2, retrieved 2026-11-17, license: not stated.

| Rank | Model | Final money balance |
|---|---|---|
| 1 | Gemini 4 Carbon | $16,484.21 ± $1,096 |
| 2 | GPT-6 Astra | $15,514.70 ± $1,074 |
| 3 | Gemini 4 Hydrogen | $14,549.49 ± $495 |
| 4 | GPT-6 Sol | $14,427.85 ± $1,051 |
| 5 | Gemini 4 Argon | $13,718.16 ± $3,100 |
| 6 | GPT-6.1 Sol | $12,860.61 ± $1,043 |
| 7 | Muse Wonder Full | $12,291.56 ± $885 |
| 8 | Claude Opus 5 | $11,181.87 ± $2,094 |
| 9 | Claude Sonnet 5.5 | $11,165.28 ± $1,585 |
| 10 | Claude Opus 4.7 | $10,936.76 ± $1,181 |

Notes: Top 10 of 76 shown on page; no update date shown, but includes very recent models.
