Excel Modeling Benchmark v1.1 Template
Professional Work · 2026-10-19
Template-mode score on Excel Modeling Benchmark v1.1, where agents complete a stripped template workbook and are graded by exact cell match against the gold workbook.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5.5 | 81.4 |
| Claude Sonnet 5.5 | 80.9 |
| Claude Fable 5.1 | 80.8 |
| Muse Wonder Full | 79.9 |
| Gemini 4 Argon | 79.6 |
| Gemini 3.8 Flash | 76.9 |
| Muse Wonder | 76.3 |
| Kimi K3 | 75.6 |
| Grok 4.7 | 75.1 |
| GPT-6.1 Sol | 75.1 |
| GPT-6 Astra | 74.8 |
| Claude Haiku 5.5 | 73.4 |
| MiMo-V2.6-Pro | 73.0 |
| Mistral Large 4 Preview | 72.7 |
| Step 5 Preview | 72.2 |
| GLM-5.3 | 71.7 |
| MiMo-V2.6-Flash | 71.7 |
| MiniMax M3.1 Flash | 71.6 |
| GPT-6 Luna | 71.4 |
| DeepSeek V4.1 Flash 1017 | 70.8 |
| Hy4 Preview | 70.5 |
| DeepSeek V4.1 Flash | 70.2 |
| GLM-5.3 Flash | 70.1 |
| Beam 1.0 501B A23B | 67.9 |
| Qwen3.8-Max | 65.3 |
| DeepSeek V4 Pro 0813 | 64.3 |
| Muse Spark Open | 61.8 |
| MiniMax M3 | 60.0 |
| Mercury 3 | 53.0 |
| Inkling | 38.4 |
| Inkling-Small | 37.8 |