NoisyBench - Random Chat History: leaderboard
Metric: Harmonic Mean Score (%; 11 datasets' pass@k, pass^k for TauBench; a random WildChat chat history added). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 62.5 |
| 2 | DeepSeek R1 0528 | 59.4 |
| 3 | Gemini 2.5 Flash | 56.9 |
| 4 | GPT-OSS-120B (High) | 54.9 |
| 5 | Qwen 3 30B A3B 2507 (Thinking) | 49.3 |
| 6 | Qwen 3 4B 2507 (Thinking) | 46.5 |
| 7 | DeepSeek R1 Distill Llama 8B | 23 |
Interactive version: theaggregate.ai/benchmark?slug=noisybench-random-chat-history · How It Works · Data refreshed daily, snapshot 2026-09-25.