# Microsoft reveals cutting-edge MAI-Base-2 model with 2T parameters

DevFeed: [Microsoft reveals cutting-edge MAI-Base-2 model with 2T parameters](<https://devfeed.tech/articles/microsoft-reveals-cutting-edge-mai-base-2-model-with-2t-parameters-88918.md>)

Original publisher: [Read original article](<https://siliconangle.com/2026/10/16/microsoft-reveals-cutting-edge-mai-base-2-model-with-2t-parameters/>)

Author: Maria Deutscher

Published: 2026-10-17T00:39:50Z

Content type: news

Language: en

Sources: [SiliconANGLE](<https://devfeed.tech/sources/siliconangle.md>)

Topics: [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Training AI Models](<https://devfeed.tech/topics/training-ai-models.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Open Source Models & Datasets](<https://devfeed.tech/topics/open-source-models-datasets.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [large-language-model](<https://devfeed.tech/tags/large-language-model.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llms](<https://devfeed.tech/tags/llms.md>), [mai-base-2](<https://devfeed.tech/tags/mai-base-2.md>), [microsoft](<https://devfeed.tech/tags/microsoft.md>), [mixture-of-experts](<https://devfeed.tech/tags/mixture-of-experts.md>), [muon](<https://devfeed.tech/tags/muon.md>), [nasdaq-msft](<https://devfeed.tech/tags/nasdaq-msft.md>), [news](<https://devfeed.tech/tags/news.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [the-latest](<https://devfeed.tech/tags/the-latest.md>)

## AI overview

Microsoft introduced MAI-Base-2, a 2-trillion-parameter base model that activates 80 billion parameters per prompt and uses a mixture-of-experts architecture. The company reports stronger results than DeepSeek-V4-Pro-Base on internal science, programming, and math benchmarks, and up to 12 times greater hardware efficiency by certain measures. Microsoft trained it on 16,000 GB200 chips and plans to license its weights to trusted partners through Azure.

## Source excerpt

Microsoft Corp. has developed a new large language model that it says is significantly more efficient than leading open-source alternatives. The company's artificial intelligence research group detailed the algorithm on Thursday. MAI-Base-2 is not a full-fledged LLM but rather a base model, a kind of no-frills prototype. Developers turn base models into LLMs by putting [...] The post Microsoft reveals cutting-edge MAI-Base-2 model with 2T parameters appeared first on SiliconANGLE.