FINESSE-Bench - CFA-like Level 3: leaderboard

FINESSE-Bench CFA-like Level 3 tests financial knowledge with 318 multiple-choice questions on portfolio management, wealth planning, risk management and ethics cases from FINESSE-Bench (2026). Answers are marked correct or incorrect by a GPT-5.2 judge against the reference; the score is accuracy in percent, higher is better.

Metric: Accuracy (%) on the 318 CFA-like Level 3 multiple-choice questions on portfolio management, wealth planning, risk management and ethics cases of FINESSE-Bench; a GPT-5.2 judge marks each answer correct or incorrect against the reference; zero-shot, one fixed prompt per task type, temperature 0 where possible, reasoning configurations with medium effort where the model offers them; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 31 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1GLM-583.96#269
2Claude Sonnet 4.6Medium82.39#173
3Qwen 3.5 Plus (2026-02-15)82.08#194
4Kimi K2.581.13#266
5GLM-4.780.19#380
6GPT-5.2Medium80.19#185
7Claude 3.7 SonnetThinking78.93#470
8DeepSeek R1 052877.36#428
9Qwen 3.5 122B A10B76.73#346
10Qwen 3.5 27B76.42#375
11Qwen 3.5 35B A3B75.16#492
12Qwen 3 235B A22B 2507 (Thinking)74.84#466
13Qwen 3.5 Flash (02-23)74.53#557
14Qwen 3.5 397B A17B72.01#264
15Llama 4 Maverick71.7#1110

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=finesse-bench-cfa-like-level-3 · How It Works · Data refreshed daily, snapshot 2026-11-21.