See something missing, incorrect, or out of date? Tell us about this page.
AI benchmark results and model performance
Individual benchmark scores plotted by date.
| Organisation | Model | Reported | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 31 Jul 2026 | 25.20% | DeepSeek-V4-Flash-0731 | Yes | Source | |
| DeepSeek V4.1 Flash | 10 Sept 2026 | 31.80% | Instruct model; reasoning_effort=100; Pass@1 | Yes | Source | |
| Qwen 3.8 27B | 14 Aug 2026 | 42.90% | Pass@1: 20.4; reported aggregate score: 42.9 | Yes | Source | |
| GPT-6 Astra | 03 Sept 2026 | 59.30% | GPT-6 Astra launch evaluation | Yes | Source |