Yandex AliceAI Cards - MMLU-Pro (CoT): leaderboard
MMLU-Pro ten-option knowledge and reasoning questions, 5-shot with chain of thought, run by Yandex. Yandex AliceAI-Foundation-80B-A3B-Base and AliceAI-T5-35B-A0.6B model cards (September 2026), joined on the runs both cards share; pretrained base checkpoints; accuracy in percent, higher is better.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around February 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 35B A3B Base | 63.2 | #319 |
| 2 | Qwen3.5-4B-Base | 51.1 | #801 |
| 3 | Gemma 4 E4B | 37.4 | #764 |
| 4 | Qwen3.5-2B-Base | 36.5 | #1288 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=yandex-aliceai-cards-mmlu-pro-cot · How It Works · Data refreshed daily, snapshot 2026-11-07.