XpertBench - Finance: leaderboard
Metric: Weighted rubric score (%) on the XpertBench-Gold subset (245 expert-written open-ended professional tasks, 15-40 weighted binary checkpoints each); ShotJudge: Gemini 2.5 Pro judges every checkpoint with an expert-graded GPT-5 response as one-shot exemplar; finance tasks; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 (High) | 84.65 |
| 2 | Claude Opus 4.6 (Thinking) | 73.25 |
| 3 | GPT-5.2 (High) | 71.22 |
| 4 | GPT-5 (High) | 64.5 |
| 5 | Gemini 3 Pro | 62.96 |
| 6 | Gemini 3.1 Pro (Preview) | 62.43 |
| 7 | Gemini 3 Flash | 57.45 |
| 8 | Kimi K2 (Thinking) | 55 |
| 9 | GLM-4.6 | 52.7 |
| 10 | DeepSeek V3.1 Terminus (Thinking) | 46.1 |
| 11 | Claude Sonnet 4.5 (Thinking) | 44.5 |
| 12 | Doubao-Seed-1.6 (Thinking) | 40.8 |
| 13 | Gemini 2.5 Pro | 30.3 |
Interactive version: theaggregate.ai/benchmark?slug=xpertbench-finance · How It Works · Data refreshed daily, snapshot 2026-10-07.