YC-Bench Bankruptcy Rate
Games · 2026-04-01
Share of YC-Bench's three seeded default-preset runs in which the simulated startup went bankrupt (funds below zero) before the one-year horizon, as counted on Collinear AI's leaderboard. Lower is better; with only three runs per model it is a coarse diagnostic outside the capability fit.
Top models (lower is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 0.0 |
| Opus 4.6 | 0.0 |
| Opus 4.7 | 0.0 |
| Opus 4.8 | 0.0 |
| Claude Opus 5 | 0.0 |
| Claude Opus 5.5 | 0.0 |
| Claude Sonnet 5.5 | 0.0 |
| DeepSeek V4.1 Flash | 0.0 |
| DeepSeek-V4-Pro | 0.0 |
| DeepSeek V4 Pro 0813 | 0.0 |
| Gemini 3.5 Flash | 0.0 |
| Gemini 3.6 Flash | 0.0 |
| Gemini 3 Flash Preview | 0.0 |
| GLM-5 | 0.0 |
| GLM-5.1 | 0.0 |
| GLM-5.2 | 0.0 |
| GLM-5.3 Flash | 0.0 |
| GLM-5.3 | 0.0 |
| GPT-5.4 | 0.0 |
| GPT-5.5 | 0.0 |
| GPT-5.6 Sol | 0.0 |
| GPT-5.6 Terra | 0.0 |
| GPT-6.1 Sol | 0.0 |
| GPT-6 Astra | 0.0 |
| GPT-6 Sol | 0.0 |
| Grok 4.5 | 0.0 |
| Grok 4.6 | 0.0 |
| Grok 4.7 | 0.0 |
| Inkling | 0.0 |
| Kimi K3 | 0.0 |
| MiMo-V2.5 | 0.0 |
| MiMo-V2.6-Pro | 0.0 |
| MiniMax M3 | 0.0 |
| Qwen3.6 Plus (2026-04-02) | 0.0 |
| Qwen3.8-Max | 0.0 |
| Qwen3.8-Omni-Flash | 0.0 |
| Sonnet 5 | 33.3 |
| Gemini 3.1 Flash-Lite Preview | 33.3 |
| Gemini 3.1 Pro Preview | 33.3 |
| GPT-5.4 Mini | 33.3 |