CEO-Bench - Max Survival Days: leaderboard
CEO-Bench longevity slice: how many simulated days an agent keeps the company solvent at its best run, rather than the cash it ends with. Reported by the benchmark authors.
Metric: Max survival days (self-reported). Source: benchmarklist.com. Saturation forecast: Around August 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.6 | 500 | #94 |
| 2 | GPT-5.5 | 500 | #33 |
| 3 | Claude Opus 4.7 | 500 | #51 |
| 4 | Kimi K2.6 | 500 | #113 |
| 5 | Claude Opus 4.8 | 500 | #39 |
| 6 | GLM-5.1 | 324 | #109 |
| 7 | Claude Haiku 4.5 | 231 | #287 |
| 8 | Gemini 3 Flash | 226 | #103 |
| 9 | DeepSeek V4 Pro | 176 | #112 |
| 10 | Grok 4.20 | 37 | #179 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ceo-bench-max-survival-days · How It Works · Data refreshed daily, snapshot 2026-11-05.