FIRE (Finance, Qualification Exams) - Banking (CCBP): leaderboard

Metric: Accuracy (%) on the Banking (CCBP) questions of FIRE's financial-qualification set (about 14,000 multiple-choice questions from 14 professional certification exams); exact option match; each model with its recommended inference settings, else temperature 0.6 and top-p 0.8; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 14 models tracked.

Top models

#ModelScoreOverall rank
1Seed-1.692.16#257
2Gemini 3 Pro90.06#77
3Qwen 3 Max89.04#201
4Seed-OSS-36B-Instruct88.05#348
5Qwen 3 235B A22B 2507 (Thinking)86.92#253 (Qwen 3 235B A22B 2507)
6Kimi K2 (Thinking)85.67#236 (Kimi K2)
7DeepSeek V3.285.35#198
8GPT-5.282.19#105
9GLM-4.681.28#246
10Claude Sonnet 4.579.21#138
11GPT-OSS-120B68.23#330

No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.

Interactive version: theaggregate.ai/benchmark?slug=fire-finance-qualification-exams-banking-ccbp · How It Works · Data refreshed daily, snapshot 2026-10-11.