Excel Modeling Benchmark Comps
Professional Work · 2026-07-01
This row reports the category score for comparable-company analysis tasks in Excel Modeling Benchmark.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5 | 75.2 |
| Opus 4.8 | 72.8 |
| Kimi K3 | 72.7 |
| GPT-5.6 Terra | 72.5 |
| GPT-5.6 Sol | 71.3 |
| Claude Fable 5 | 70.8 |
| Gemini 3.6 Flash | 70.0 |
| Gemini 3.5 Flash | 69.8 |
| Sonnet 5 | 69.4 |
| GPT-5.5 | 64.7 |
| Sonnet 4.6 | 61.5 |
| Kimi K2.6 | 61.4 |
| DeepSeek-V4-Pro | 57.4 |
| GLM-5.2 | 54.9 |
| GPT-5.6 Luna | 54.6 |
| MiniMax M3 | 52.8 |
| Muse Spark 1.1 | 52.7 |
| Qwen3.7-Max | 50.9 |
| MiMo-V2.5-Pro | 47.3 |
| GPT-5.4 Nano | 46.9 |
| Gemini 3.5 Flash-Lite | 46.4 |
| Gemini 3.1 Pro Preview | 44.5 |
| GPT-5.4 Mini | 40.5 |
| Grok 4.5 | 39.5 |
| Qwen3.7-Plus | 39.1 |
| Inkling | 33.0 |
| Inkling-Small | 30.8 |
| Grok 4.3 | 16.6 |
| Gemini 3.1 Flash-Lite Preview | 10.5 |