Atlas

Benchmarks

← All benchmarks

Vending-Bench 2

Games · 2025-11-18

Final bank-balance score in a simulated vending business benchmark.

Top models (higher is better)

ModelScore
Gemini 4 Carbon16484
GPT-6 Astra15515
GPT-6 Sol14428
Gemini 4 Argon13718
Muse Wonder Full12292
Claude Opus 511182
Claude Sonnet 5.511165
Opus 4.710937
Grok 4.710537
GPT-5.6 Sol9619
Muse Wonder9432
Claude Opus 5.59235
Grok 4.69047
GLM-5.28314
GLM-5.38164
Opus 4.68018
GPT-5.57524
GPT-5.6 Terra7343
Sonnet 4.67204
Sonnet 56378
Kimi K2.66205
GPT-5.46144
GPT-5.3-Codex5940
Opus 4.85787
Claude Fable 55680
GLM-5.15634
Gemini 3 Pro Preview5478
Claude Fable 5.15422
Gemini 3.5 Flash5396
Qwen3.6 Plus (2026-04-02)5115
Gemini 3.8 Flash5094
Kimi K2.7 Code5083
Opus 4.54967
Grok 4.204663
GLM-54432
Qwen3.6-Max-Preview4254
GPT-5.6 Luna4095
GPT-5.23952
Grok 4.53887
Sonnet 4.53839
Loading Atlas data…