Excel Modeling Benchmark v1.1 Comps
Professional Work · 2026-10-19
Category score for comparable-company analysis tasks in Excel Modeling Benchmark v1.1.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 82.9 |
| Claude Fable 5.1 | 82.0 |
| Muse Wonder Full | 80.2 |
| Claude Sonnet 5.5 | 79.4 |
| Muse Wonder | 77.9 |
| Claude Haiku 5.5 | 77.7 |
| GPT-6 Astra | 76.5 |
| Kimi K3 | 76.4 |
| Gemini 3.8 Flash | 76.0 |
| Claude Opus 5.5 | 75.7 |
| Grok 4.7 | 72.4 |
| MiniMax M3.1 Flash | 71.6 |
| DeepSeek V4.1 Flash 1017 | 70.8 |
| Mistral Large 4 Preview | 70.5 |
| GPT-6 Luna | 69.7 |
| MiMo-V2.6-Flash | 69.3 |
| Hy4 Preview | 68.5 |
| DeepSeek V4.1 Flash | 68.0 |
| MiMo-V2.6-Pro | 68.0 |
| GLM-5.3 Flash | 67.7 |
| GLM-5.3 | 67.6 |
| Step 5 Preview | 66.8 |
| Qwen3.8-Max | 65.7 |
| Beam 1.0 501B A23B | 63.0 |
| GPT-6.1 Sol | 61.2 |
| MiniMax M3 | 60.1 |
| DeepSeek V4 Pro 0813 | 59.3 |
| Mercury 3 | 59.3 |
| Muse Spark Open | 56.0 |
| Inkling-Small | 45.6 |
| Inkling | 40.7 |