ReLE - Finance: leaderboard
Metric: Accuracy (%). Source: nonelinear.com. 178 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K3 | 89.7 |
| 2 | Qwen 3.7 Max | 89.1 |
| 3 | Seed 2.0 Pro | 87.7 |
| 4 | Kimi K2.5 (Thinking) | 87.6 |
| 5 | Kimi K2.7 Code | 87 |
| 6 | Qwen 3.8 Max | 86.9 |
| 7 | Qwen 3.6 Max Preview | 86.5 |
| 8 | Kimi K2.6 | 86.4 |
| 9 | Qwen 3.7 Plus | 85.7 |
| 10 | GPT-5.6 Pro Sol | 85.7 |
| 11 | MiMo-V2.5-Pro | 85.6 |
| 12 | GLM-5.1 | 85.2 |
| 13 | GPT-5.4 (High) | 85.1 |
| 14 | Qwen 3 Max (Preview) | 85.1 |
| 15 | DeepSeek V4 Pro | 84.6 |
Interactive version: theaggregate.ai/benchmark?slug=rele-finance · How It Works · Data refreshed daily, snapshot 2026-09-19.