---
title: "The verification node for Microsoft's transformation is approaching."
type: "Topics"
locale: "en"
url: "https://longbridge.com/en/topics/41297941.md"
description: "I used to think Microsoft's self-developed models were too slow and couldn't compare to the big three. But the upcoming AI PC shows me that the foresight of Microsoft's strategy is about to be revealed. Although the performance of its self-developed models can't match Claude, ChatGPT, or Gemini, the small self-developed models can run locally, have lower costs for enterprise use, can meet small daily work demands, and are more suitable for enterprise work scenarios. Moreover, local model data stays local without uploading to the cloud, ensuring the privacy of enterprise information, making companies more willing to use them..."
datetime: "2026-05-30T15:30:13.000Z"
locales:
  - [en](https://longbridge.com/en/topics/41297941.md)
  - [zh-CN](https://longbridge.com/zh-CN/topics/41297941.md)
  - [zh-HK](https://longbridge.com/zh-HK/topics/41297941.md)
author: "[左侧+耐心](https://longbridge.com/en/profiles/18725498.md)"
---

# The verification node for Microsoft's transformation is approaching.

I used to think Microsoft's self-developed models were slow to develop, not even comparable to the top three.

But the upcoming AI PC has shown me that Microsoft's forward-looking layout is beginning to emerge. Although its self-developed models are not as powerful as Claude, ChatGPT, and Gemini in terms of performance, the small self-developed models can run locally, with low costs for enterprise use, meeting small daily work needs, and fitting better into enterprise work scenarios. Moreover, local model data is stored locally without needing to be uploaded to the cloud, ensuring the privacy of enterprise information, making enterprises more willing to use them.

Once Microsoft's operating system in the AI era can automatically judge task complexity, deciding whether to use a local small model or a cloud large model to complete tasks, and automatically invoke them, then Microsoft will successfully control the era's entry point for personal AI use. The top three will instead become options, not the underlying system, and Copilot will also become the enterprise's first choice. After all, it comes with the computer system and is low-cost, better meeting enterprise production needs. Microsoft's growth explosion point is coming. Even without considering Copilot's profit growth, it's close to 20%, with a PE of only 27 (Microsoft's average PE as a software stock is around 33, with a peak of 40). If priced as an AI distribution entry point, how much should its growth be? How high should its PE go? 🤔  

Seeing 600 by year-end

$Microsoft(MSFT.US)

### Related Stocks

- [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md)
- [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md)
- [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md)
- [CHA.US](https://longbridge.com/en/quote/CHA.US.md)
- [MSFL.US](https://longbridge.com/en/quote/MSFL.US.md)
- [MSFO.US](https://longbridge.com/en/quote/MSFO.US.md)
- [MSFD.US](https://longbridge.com/en/quote/MSFD.US.md)
- [MSFU.US](https://longbridge.com/en/quote/MSFU.US.md)
- [MSFX.US](https://longbridge.com/en/quote/MSFX.US.md)
- [MSFY.US](https://longbridge.com/en/quote/MSFY.US.md)
- [GGLS.US](https://longbridge.com/en/quote/GGLS.US.md)
- [GGLL.US](https://longbridge.com/en/quote/GGLL.US.md)