UnpredictaBench - Code: leaderboard
Metric: KS@100 (%): share of the 238 Code problems (distributions induced by stochastic programs) for which 100 samples from the model (100 independent calls at temperature 1.0, reasoning effort none, up to 5 retries for a parseable answer) are not rejected by a two-sample Kolmogorov-Smirnov test against ground-truth samples at p < 0.0001; higher is better. Source: arxiv.org. Saturation forecast: Around 2040. 20 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o | 29.41 |
| 2 | Nemotron 3 Nano 30B A3B (Non-reasoning) | 18.49 |
| 3 | Ministral-3-3B-Instruct-2512 | 17.23 |
| 4 | DeepSeek V3.2 (Non-reasoning) | 14.29 |
| 5 | Qwen 3.5 2B (Non-reasoning) | 12.61 |
| 6 | GPT-4o Mini | 10.5 |
| 7 | GPT-5.4 (Non-reasoning) | 10.5 |
| 8 | Llama 3.1 70B Instruct | 9.66 |
| 9 | OLMo 3 7B Instruct | 5.46 |
| 10 | Qwen 3 32B (Non-reasoning) | 5.46 |
| 11 | Grok 4.1 Fast (Non-reasoning) | 5.46 |
| 12 | Claude Sonnet 4.6 | 5.04 |
| 13 | Mistral Large 3 | 5.04 |
| 14 | Llama 3.1 8B Instruct | 4.62 |
| 15 | Llama 3.2 1B Instruct | 4.2 |
Interactive version: theaggregate.ai/benchmark?slug=unpredictabench-code · How It Works · Data refreshed daily, snapshot 2026-09-29.