YandexGPT 5.1 Pro: benchmark results

Provider: Yandex. Access: API.

Aggregate Rating 1529 ± 29, rank #1957 of 10916 rated models, from 13 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
LLM Arena RU990Arena Elo44.6
FINESSE-Bench - CMT-like Level 277.69Accuracy (%) on the 251 CMT-like Level 2 multiple-choice que30
FINESSE-Bench - CFA-like Level 360.06Accuracy (%) on the 318 CFA-like Level 3 multiple-choice que26.7
FINESSE-Bench - CFA-like Level 253.58Accuracy (%) on the 293 CFA-like Level 2 multiple-choice ite20
FINESSE-Bench - VLigaBench-ru55.56Accuracy (%) on the 324 VLigaBench-ru Russian-language olymp20
FINESSE-Bench64.26exam-like (self-reported)17.2
FINESSE-Bench - CFA-like Level 167.91Accuracy (%) on the 1069 CFA-like Level 1 multiple-choice qu16.7
FINESSE-Bench - CFTe-like Level 173.37Accuracy (%) on the 781 CFTe-like Level 1 multiple-choice qu16.7
FINESSE-Bench - Trading Derivatives44.85Accuracy (%) on the 544 Trading_derivatives numerical and sh16.7
FINESSE-Bench - Trading TA66.83Accuracy (%) on the 413 Trading_TA numerical and short-answe16.7
FINESSE-Bench - Trading and Technical Analysis62.89Accuracy (%) over the trading and technical-analysis group o16.7
PRISM (1C:Enterprise) - Platform Tasks (B)101C platform tasks fully solved, run in headless 1C (%)12.5

Interactive version: theaggregate.ai/model?slug=yandexgpt-5-1-pro · How It Works · Data refreshed daily, snapshot 2026-11-08.