FINESSE-Bench - Trading TA: leaderboard

FINESSE-Bench Trading TA tests financial knowledge with 413 numerical and short-answer applied technical-analysis tasks (patterns, momentum and mean reversion, entries and exits, backtesting) 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 413 Trading_TA numerical and short-answer applied technical-analysis tasks (patterns, momentum and mean reversion, entries and exits, backtesting) 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
1GPT-5.2Medium83.54#185
2Claude Sonnet 4.6Medium83.54#173
3Kimi K2.583.29#266
4GLM-582.57#269
5Qwen 3.5 Plus (2026-02-15)82.32#194
6Qwen 3.5 397B A17B81.84#264
7GLM-4.781.6#380
8GPT-5.4Medium80.87#135
9Qwen 3.5 27B79.66#375
10Qwen 3.5 35B A3B78.69#492
11Qwen 3.5 122B A10B78.21#346
12Claude 3.7 SonnetThinking77.97#470
13Qwen 3 235B A22B 2507 (Thinking)77.72#466
14Qwen 3.5 Flash (02-23)77.24#557
15DeepSeek R1 052876.51#428

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-trading-ta · How It Works · Data refreshed daily, snapshot 2026-11-21.