Finance Agent v2 - Adjustments
Professional Work · 2026-05-19
The Adjustments split of Finance Agent v2. This child benchmark separates a source-reported subtask or subtrack from the parent aggregate so scores at different grains do not share one benchmark_slug.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 60.5 |
| DeepSeek V4.1 Flash 1017 | 58.4 |
| Muse Spark Open | 57.2 |
| Muse Wonder Full | 56.7 |
| Muse Spark 1.2 | 56.3 |
| Qwen3.8 Max (0902) | 55.4 |
| Gemini 3.5 Flash | 54.9 |
| Gemini 3.8 Flash | 54.8 |
| Gemini 3.7 Flash | 54.7 |
| Claude Fable 5.1 | 54.4 |
| Gemini 3.6 Flash | 53.6 |
| GLM-5.3 Flash | 53.2 |
| Muse Spark 1.3 | 52.8 |
| GLM-5.3 | 52.7 |
| Muse Wonder | 52.7 |
| Mistral Large 4 Preview | 52.4 |
| GPT-5.6 Luna | 52.0 |
| MiMo-V2.6-Pro | 51.6 |
| MiMo-V2.6-Flash | 51.4 |
| Ember-1 | 51.3 |
| GPT-6 Luna | 49.8 |
| GPT-5.6 Terra | 49.7 |
| Muse Spark 1.1 | 49.7 |
| Grok 4.6 | 49.6 |
| Claude Haiku 5.5 | 49.6 |
| DeepSeek V4.1 Flash | 49.2 |
| GPT-5.6 Sol | 49.2 |
| Sonnet 4.6 | 48.9 |
| DeepSeek V4 Pro 0813 | 48.8 |
| Hy4 Preview | 48.8 |
| Sonnet 5 | 48.7 |
| Qwen3.8-Max | 48.4 |
| Opus 4.8 | 48.4 |
| GPT-6 Astra | 48.3 |
| DeepSeek-V4-Flash-0731 | 47.8 |
| Step 5 Preview | 47.8 |
| Qwen3.8 27B | 47.3 |
| GLM-5.2 | 46.3 |
| Grok 4.7 | 46.3 |
| GPT-5.5 | 45.7 |