| Qwen3.8-Omni-Flash18 Sep 2026 · Alibaba | 81.5% | Qwen3.8-Omni-Flash: Omni Senses. Agentic Delivery.Our capture | |
| Qwen3.8-Flash-Next26 Aug 2026 · Alibaba | 81.3% | Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-EfficiencyOur capture | CorroboratedThe same number is printed in 1 other lab document: 1 |
| Qwen3.8-Max3 Aug 2026 · Alibaba | 82.8% | Qwen3.8-Max: A New Bar for Coding and CoworkOur capture | |
| Inkling15 Jul 2026 · Thinking Machines | 79.8% | Inkling: Our Open-Weights Modeleffort 0.99Our capture | CorroboratedThe same number is printed in 2 other lab documents: 1 2 |
| Qwen3.7-Plus1 Jun 2026 · Alibaba | 79.1% | Qwen3.7-Plus: Multimodal Agent IntelligenceOur capture | CorroboratedThe same number is printed in 3 other lab documents: 1 2 3 |
| Mistral Medium 3.522 May 2026 · Mistral | 69.0% | Remote agents in Vibe. Powered by Mistral Medium 3.5.maximum reasoningSource image | |
| Qwen3.7-Max20 May 2026 · Alibaba | 79.1% | Qwen3.7: The Agent Frontier | CorroboratedThe same number is printed in 1 other lab document: 1 |
| Qwen3.6-Plus2 Apr 2026 · Alibaba | 74.2% | Qwen3.6-Plus: Towards Real World Agents | CorroboratedThe same number is printed in 2 other lab documents: 1 2 |
| Qwen3.5-Omni30 Mar 2026 · Alibaba | 52.6% | Qwen3.5-Omni: Scaling Up, Toward Native Omni-Modal AGIQwen3.5-Omni-PlusOur capture | |
| Qwen3.5-Omni-Flash30 Mar 2026 · Alibaba | 38.4% | Qwen3.5-Omni: Scaling Up, Toward Native Omni-Modal AGIQwen3.5-Omni-FlashOur capture | |
| Mistral Small 4 119B16 Mar 2026 · Mistral | 48% | Introducing Mistral Small 4reasoning modeSource image | |
| Mistral Small 4 119B16 Mar 2026 · Mistral | 35.7% | Introducing Mistral Small 4instruct mode, no reasoning; another report, not chartedSource image | |
| Qwen3.5-397B-A17B16 Feb 2026 · Alibaba | 76.5% | Qwen3.5: Towards Native Multimodal Agents | CorroboratedThe same number is printed in 3 other lab documents: 1 2 3 |
| MiniMax M2.512 Feb 2026 · MiniMax | 70.0% | MiniMax M2.5: Built for Real-World Productivity.internal run following Artificial Analysis Intelligence Index methodsSource image | |
| Qwen3-Max-Thinking26 Jan 2026 · Alibaba | 70.9% | Pushing Qwen3-Max-Thinking Beyond its LimitsOur capture | CorroboratedThe same number is printed in 2 other lab documents: 1 2 |
| MiniMax M2.123 Dec 2025 · MiniMax | 70.0% | MiniMax M2.1: Significantly Enhanced Multi-Language Programming, Built for Real-World Complex Tasksinternal run following Artificial Analysis Intelligence Index methodsSource image | CorroboratedThe same number is printed in 3 other lab documents: 1 2 3 |
| Mistral Large 3 675B2 Dec 2025 · Mistral | 37.8% | Introducing Mistral Small 4instruct model; reported laterSource image | |
| MiniMax M227 Oct 2025 · MiniMax | 72.0% | MiniMax M2.1: Significantly Enhanced Multi-Language Programming, Built for Real-World Complex Tasksinternal run following Artificial Analysis Intelligence Index methods; reported laterSource image | CorroboratedThe same number is printed in 2 other lab documents: 1 2 |
