Excel Modeling Benchmark (EMB)
As of Nov 3, 2026, Excel Modeling Benchmark (EMB) (v1.1) is maintained by Vals AI. Metric accuracy (%). Ceiling 100. Saturation unknown. Contamination risk low.
- Maintainer
- Vals AI
- Metric
- accuracy (%)
- Ceiling
- 100
- Saturation
- unknown
- Contamination
- low
- Cadence
- Declared cadence: as Vals updates (checked daily). Catalog snapshots: 1, latest Nov 3, 2026.
- Source
- Vals AI · this benchmark on Vals AI
- Grading
- Mixed (exact checks plus LLM-graded rubric) — Template mode is exact cell match; scratch mode is a partial-credit rubric checked by an LLM judge.
Source: Vals AI. A proprietary private test set run by Vals; scores are republished with permission. Rows are dated the day isitsota first saw them, and only the benchmark's standard harness counts.
Categories
Where it counts
- Finance — secondary, weight 0.35
Top 5
Current top scores
- Claude Fable 5.1 · 76.728 · Nov 3, 2026
- Gemini 4 Carbon · 76.468 · Nov 3, 2026
- Claude Sonnet 5.5 · 75.899 · Nov 3, 2026
- Claude Opus 5.5 · 75.836 · Nov 3, 2026
- Muse Wonder Full · 74.998 · Nov 3, 2026
| Model | Score | Date | Kind |
|---|---|---|---|
| Claude Fable 5.1 | 76.728 | 2026-11-03 | snapshot |
| Claude Sonnet 5.5 | 75.899 | 2026-11-03 | snapshot |
| Claude Opus 5.5 | 75.836 | 2026-11-03 | snapshot |
| Muse Wonder Full | 74.998 | 2026-11-03 | snapshot |
| Gemini 4 Argon | 74.828 | 2026-11-03 | snapshot |
| Gemini 3.8 Flash | 72.871 | 2026-11-03 | snapshot |
| GPT-6 Astra | 72.683 | 2026-11-03 | snapshot |
| Muse Wonder | 72.503 | 2026-11-03 | snapshot |
| Grok 4.7 | 71.498 | 2026-11-03 | snapshot |
| Kimi K3 | 71.264 | 2026-11-03 | snapshot |
| GPT-6.1 Sol | 71.088 | 2026-11-03 | snapshot |
| GPT-6 Luna | 69.954 | 2026-11-03 | snapshot |
| Step 5 Preview | 69.552 | 2026-11-03 | snapshot |
| DeepSeek V4.1 Flash 1017 | 69.173 | 2026-11-03 | snapshot |
| GLM 5.3 | 69.112 | 2026-11-03 | snapshot |
| Mistral Large 4 | 68.972 | 2026-11-03 | snapshot |
| Hy4 Preview | 68.946 | 2026-11-03 | snapshot |
| Mimo V2.6 Flash | 68.882 | 2026-11-03 | snapshot |
| MiniMax M3.1 Flash | 67.843 | 2026-11-03 | snapshot |
| DeepSeek V4.1 Flash | 67.758 | 2026-11-03 | snapshot |
| Mimo V2.6 Pro | 67.347 | 2026-11-03 | snapshot |
| GLM 5.3 Flash | 66.252 | 2026-11-03 | snapshot |
| Qwen3 8 Max | 65.986 | 2026-11-03 | snapshot |
| Claude Haiku 5.5 | 65.477 | 2026-11-03 | snapshot |
| DeepSeek V4 Pro 0813 | 65.128 | 2026-11-03 | snapshot |
| Muse Spark Open | 61.983 | 2026-11-03 | snapshot |
| Minimax M3 | 59.754 | 2026-11-03 | snapshot |
| Inkling | 40.971 | 2026-11-03 | snapshot |
| Inkling Small | 40.823 | 2026-11-03 | snapshot |
| Mercury 3 | 57.88 | 2026-11-03 | snapshot |