AA-Omniscience Business Accuracy
General QA
AA-Omniscience accuracy on its business domain (accounting, corporate finance, economics and investments), as published in the Artificial Analysis Finance & Accounting Index. A subset of the AA-Omniscience v1.1 questions.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-6.1 Sol | 53.8 |
| GPT-6 Astra | 52.4 |
| Gemini 4 Carbon | 46.4 |
| Gemini 3.8 Flash | 43.8 |
| Grok 4.7 | 39.3 |
| MiniMax M3.1 Flash | 39.1 |
| Inkling | 38.7 |
| Kimi K3.1 Pro | 38.7 |
| GPT-6.1 Terra | 38.2 |
| Claude Sonnet 5.5 | 37.8 |
| Muse Spark Open | 35.6 |
| GPT-6 Luna | 34.2 |
| Muse Wonder Full | 31.8 |
| Step 5 Preview | 30.4 |
| MiMo-V2.6-Pro | 29.8 |
| DeepSeek V4.1 Flash 1017 | 29.3 |
| Muse Wonder | 28.0 |
| Claude Haiku 5.5 | 25.6 |
| GLM-5.3 | 23.6 |
| Beam 1.0 501B A23B | 23.3 |
| Qwen3.8 Max (0902) | 21.1 |
| Ling 3.1 Flash 1019 | 20.4 |
| Mistral Large 4 | 20.2 |
| Nemotron 3.5 Super VL 120B A12B (precision unspecified) | 17.1 |
| Qwen3.8 27B | 13.3 |
| K2 Horizon 375B A23B | 11.1 |