Shlepa - Law MC (RU): leaderboard

The Law multiple-choice track of Shlepa — Russian-language legal knowledge and reasoning questions; accuracy (%).

Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 18 models tracked.

Top models

#ModelScoreOverall rank
1Gemma 2 9B58.19#958
2Phi-3-medium-4k-instruct54.07#952
3suzume-llama-3-8B-multilingual51.8#1084
4Starling-LM-7B-beta48.51#1093
5Llama 3 8B SFR Iterative DPO R47.17#1109
6Llama 2 7B Base11.43#1518

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=shlepa-law-mc-ru · How It Works · Data refreshed daily, snapshot 2026-10-20.