Finance Agent v2.1 - Comparables
Professional Work · 2026-10-19
Partial-credit score on the Comparables category of Finance Agent v2.1.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 61.7 |
| DeepSeek V4.1 Flash 1017 | 59.8 |
| Gemini 3.8 Flash | 59.2 |
| Kimi K3 | 59.2 |
| MiMo-V2.6-Pro | 58.6 |
| GLM-5.3 | 58.5 |
| Mistral Large 4 Preview | 57.7 |
| MiMo-V2.6-Flash | 57.2 |
| Claude Haiku 5.5 | 57.0 |
| GPT-6 Luna | 56.7 |
| DeepSeek V4.1 Flash | 55.8 |
| Muse Wonder Full | 55.0 |
| Muse Wonder | 55.0 |
| Grok 4.7 | 54.9 |
| DeepSeek V4 Pro 0813 | 54.5 |
| Muse Spark Open | 54.4 |
| Step 5 Preview | 53.6 |
| Hy4 Preview | 53.3 |
| GLM-5.3 Flash | 53.1 |
| MiniMax M3.1 Flash | 53.0 |
| GPT-6.1 Sol | 50.1 |
| GPT-6 Astra | 50.1 |
| MiniMax M3 | 48.5 |
| Beam 1.0 501B A23B | 46.7 |
| Inkling | 45.8 |
| Mercury 3 | 44.9 |
| Inkling-Small | 42.5 |