AI benchmark results and model performance
Individual benchmark scores plotted by date.
| Organisation | Model | Reported | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|---|
| Gemma 3 27B | 12 Mar 2025 | 40.26 | Confab 66.3%; non-response 14.2% | No | Source | |
| GPT 4o Mini (2024-07-18) | 18 Jul 2024 | 37.21 | Confab 60.9%; non-response 13.5% | No | Source | |
| Claude 3.5 Haiku | 04 Nov 2024 | 36.74 | Confab 65.8%; non-response 7.6% | No | Source | |
| Claude 3 Haiku | 13 Mar 2024 | 34.21 | Confab 56.9%; non-response 11.5% | No | Source | |
| Nova Pro 1.0 | 04 Dec 2024 | 30.05 | Confab 54.5%; non-response 5.6% | No | Source | |
| Phi 4 | 12 Dec 2024 | 29.43 | Confab 52.5%; non-response 6.4% | No | Source | |
| GPT 4 Turbo (2024-04-09) | 09 Apr 2024 | 28.42 | Confab 26.7%; non-response 30.1% | No | Source | |
| Gemma 2 27B | 27 Jun 2024 | 27.12 | Confab 47.0%; non-response 7.2% | No | Source | |
| Gemini 2.0 Flash | 05 Feb 2025 | 26.85 | Confab 24.3%; non-response 29.4% | No | Source | |
| DeepSeek V3 (2025-03-24) | 24 Mar 2025 | 26.15 | Confab 39.1%; non-response 13.2% | No | Source | |
| Mistral Small 3.0 | 30 Jan 2025 | 25.21 | Confab 38.6%; non-response 11.8% | No | Source | |
| Minimax Text 01 | 15 Jan 2025 | 23.92 | Confab 44.6%; non-response 3.3% | No | Source | |
| Llama 3.3 70B Instruct | 06 Dec 2024 | 22.81 | Confab 17.8%; non-response 27.8% | No | Source | |
| Claude 3 Opus | 04 Mar 2024 | 22.70 | Confab 28.2%; non-response 17.2% | No | Source | |
| Llama 4 Maverick | 05 Apr 2025 | 22.58 | Confab 28.2%; non-response 16.9% | No | Source | |
| Mistral Medium 3.0 | 07 May 2025 | 21.90 | Confab 38.1%; non-response 5.7% | No | Source | |
| Mistral Large 2.0 | 24 Jul 2024 | 21.40 | Confab 32.2%; non-response 10.6% | No | Source | |
| Kimi K2 (2025-09-05) | 05 Sept 2025 | 20.38 | Confab 30.2%; non-response 10.6% | No | Source | |
| Grok 2 | 13 Aug 2024 | 20.14 | Confab 25.7%; non-response 14.5% | No | Source | |
| Claude 3.5 Sonnet (2024-10-22) | 22 Oct 2024 | 19.94 | Confab 12.9%; non-response 27.0% | No | Source | |
| Claude 3.7 Sonnet | 24 Feb 2025 | 19.76 | Confab 25.2%; non-response 14.3% | No | Source | |
| Qwen 2.5 72B | - | 19.09 | Confab 32.2%; non-response 6.0% | No | Source | |
| o1 mini | 12 Sept 2024 | 18.55 | Confab 26.2%; non-response 10.9% | No | Source | |
| Claude Opus 4.1 | 05 Aug 2025 | 18.51 | No reasoning; Confab 4.0%; non-response 33.1% | No | Source | |
| Gemini 2.0 Pro Exp (2025-02-05) | 05 Feb 2025 | 18.43 | Confab 15.8%; non-response 21.0% | No | Source |
