Finance Agent v2 - Comparables
Professional Work · 2026-05-19
The Comparables 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 3.5 Flash | 48.8 |
| GPT-5.6 Luna | 48.3 |
| Gemini 3.6 Flash | 47.8 |
| Claude Opus 5 | 47.4 |
| Muse Spark 1.1 | 47.2 |
| Kimi K3 | 46.7 |
| Sonnet 5 | 45.8 |
| Opus 4.7 | 45.6 |
| GPT-5.6 Sol | 45.3 |
| GPT-5.6 Terra | 45.2 |
| Gemini 3.5 Flash-Lite | 45.2 |
| Opus 4.8 | 44.9 |
| GLM-5.2 | 44.8 |
| Claude Fable 5 | 44.0 |
| Qwen3.6 Plus (2026-04-02) | 43.7 |
| GPT-5.5 | 43.7 |
| Inkling | 42.8 |
| Qwen3.7-Max | 42.3 |
| Grok 4.5 | 41.8 |
| MiniMax M3 | 41.7 |
| GLM-5.1 | 41.6 |
| Kimi K2.6 | 41.1 |
| Sonnet 4.6 | 41.0 |
| GPT-5.4 Mini | 40.3 |
| MiMo-V2.5-Pro | 40.1 |
| DeepSeek-V4-Pro | 39.6 |
| Qwen3.7-Plus | 38.8 |
| Gemini 3 Flash Preview | 38.6 |
| Nemotron 3 Ultra 550B A55B | 37.2 |
| GPT-5.4 Nano | 37.2 |
| Gemini 3.1 Pro Preview | 36.7 |
| Grok 4.3 | 35.9 |
| Gemini 3.1 Flash-Lite Preview | 34.1 |
| MiMo-V2.5 | 31.7 |
| Mistral Medium 3.5 | 31.5 |
| Haiku 4.5 | 31.5 |
| Grok 4.20 | 30.2 |
| MiniMax M2.7 | 28.4 |
| Laguna M.1 | 28.0 |
| Laguna XS.2 | 16.4 |