Yandex open-sources AliceAI-Foundation-80B-A3B-Base model

Yandex has released AliceAI-Foundation-80B-A3B-Base on Hugging Face under Apache 2.0, a base MoE model trained from scratch on ~18T tokens. It has 80B total and 3B active parameters, 512 experts and a 262,144-token context, and reportedly beats the previous 235B Alice AI LLM on facts, math, code and long context.

• 80B total, 3B active parameters, 512 experts, top-10 routing
• 262,144-token context, trained from scratch on ~18T tokens
• MATH-500 score 91.1 vs 81.9 for Qwen3.5 35B-A3B
• Apache 2.0 license, runs via transformers and vLLM on Nvidia GPUs