Finance Agent v2.1 - Precedents
Professional Work · 2026-10-19
Partial-credit score on the Precedents category of Finance Agent v2.1.
Top models (higher is better)
| Model | Score |
|---|---|
| MiMo-V2.6-Flash | 49.3 |
| Muse Wonder | 45.6 |
| Muse Wonder Full | 44.2 |
| Claude Haiku 5.5 | 43.1 |
| GPT-6.1 Sol | 42.7 |
| Gemini 4 Argon | 40.4 |
| GPT-6 Astra | 40.2 |
| Hy4 Preview | 39.3 |
| Muse Spark Open | 39.3 |
| MiMo-V2.6-Pro | 39.1 |
| Gemini 3.8 Flash | 38.9 |
| Mistral Large 4 Preview | 38.9 |
| GLM-5.3 Flash | 38.6 |
| DeepSeek V4.1 Flash | 37.9 |
| GPT-6 Luna | 37.0 |
| GLM-5.3 | 36.7 |
| Grok 4.7 | 35.4 |
| MiniMax M3.1 Flash | 33.9 |
| Kimi K3 | 32.6 |
| Step 5 Preview | 31.1 |
| DeepSeek V4 Pro 0813 | 30.7 |
| Beam 1.0 501B A23B | 30.6 |
| Inkling | 30.2 |
| DeepSeek V4.1 Flash 1017 | 29.3 |
| Mercury 3 | 25.0 |
| MiniMax M3 | 18.6 |
| Inkling-Small | 14.2 |