# AliceAI-Foundation-80B-A3B-Base ![Organization](https://img.shields.io/badge/Org-Yandex-FC3F1D) ![License](https://img.shields.io/badge/License-Apache%202.0-2da44e) ![Params](https://img.shields.io/badge/Params-80B%20(3B%20active)-8957e5) ![Context](https://img.shields.io/badge/Context-262K%20tokens-blue) ![Type](https://img.shields.io/badge/Type-Base%20model-555) > **Quick answer:** AliceAI-Foundation-80B-A3B-Base is Yandex's open pre-trained (base) language model: 80B total and 3B active parameters, a hybrid KDA/gated-attention MoE with 512 experts, a 262,144-token context, trained entirely from scratch and released September 12, 2026 under Apache 2.0. Yandex reports particularly strong Russian factual knowledge (WikiWebFacts 86.5, HardMultiQA 67.9) and 91.1 on MATH-500. ## At a Glance **Where it leads:** Russian-language factual knowledge and exams (WikiWebFacts 86.5, CultCat 86.5, EGE CoT 90.5) and math (MATH-500 91.1) among the open base models Yandex compared. **Where it lags:** English trivia (TriviaQA 79.0 vs 89.8 for Nemotron-3-Super-120B-A12B-Base) and SuperGPQA (44.3 vs 46.6). **Best for:** Research and fine-tuning for Russian-language assistants; it is a base model with no post-training or alignment. ## What AliceAI-Foundation-80B-A3B-Base Is Yandex built a new training corpus, chose the architecture and hyperparameters through a series of 2-trillion-token from-scratch runs, and prepared data for complex reasoning and tool use. The 48-layer network repeats a pattern of three KDA-MoE layers followed by one gated-attention-MoE layer, with 512 experts (top-10 routed plus one shared) and a 1-layer MTP head. The model is a pre-trained checkpoint intended for research, experimentation and further tuning rather than direct use in products. Alongside the weights Yandex published two Russian factual benchmarks, [WikiWebFacts](/benchmarks/yandex/wikiwebfacts) and [HardMultiQA](/benchmarks/yandex/hardmultiqa), with their evaluation protocols. ## Specifications | Field | Value | |---|---| | Organization | Yandex | | Hugging Face | [yandex/AliceAI-Foundation-80B-A3B-Base](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | | Architecture | Hybrid KDA + gated-attention MoE, 80B total / 3B active | | Languages | Russian, English | | Context length | 262,144 tokens | | License | Apache 2.0 | | Release date | 2026-09-12 | ## Public Benchmark Scores | Benchmark | Score | Source | Date | |---|---|---|---| | [WikiWebFacts](/benchmarks/yandex/wikiwebfacts) | 86.5 | [Model card](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | 2026-09 | | [HardMultiQA](/benchmarks/yandex/hardmultiqa) | 67.9 | [Model card](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | 2026-09 | | [MMLU-Pro](/benchmarks/tiger-ai-lab-university-of-waterloo/mmlu-pro) (5-shot CoT, base) | 66.8 | [Model card](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | 2026-09 | | [SuperGPQA](/benchmarks/supergpqa-team/supergpqa) (5-shot CoT, base) | 44.3 | [Model card](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | 2026-09 | | [MATH-500](/benchmarks/lightman-et-al-openai/math-500) (5-shot, base) | 91.1 | [Model card](https://huggingface.co/yandex/AliceAI-Foundation-80B-A3B-Base) | 2026-09 | *Self-reported by Yandex on a pre-trained base model; not independent Benchgen measurements. Also reported but not added: Yandex-internal benchmarks (CultCat, EduBench, ExpertFactsQA, EGE CoT, FinQA 128k, LongMemEval 128k), TriviaQA (LLM-as-judge instead of exact match), BigCodeBench (Yandex's own implementation with improved tests), LiveCodeBench v5-6 (not v6) and pass@k reasoning results (AIME 2026 pass@32 96.7, HMMT 2026 Feb pass@32 96.9, IMO AnswerBench pass@8 88.7).* ## Frequently Asked Questions

What is AliceAI-Foundation-80B-A3B-Base?

Yandex's open base language model with 80B total and 3B active parameters, trained from scratch and released September 12, 2026 under Apache 2.0.

Is it an instruction-tuned chat model?

No. It is a pre-trained base model with no post-training or alignment; Yandex recommends fine-tuning it before production use.

What is its context length?

262,144 tokens.

AliceAI-Foundation-80B-A3B-Base — Training jobs

Training jobs