Home / Technology / Microsoft Races to Cut AI Costs With New Models
Microsoft Races to Cut AI Costs With New Models
4 Jun
Summary
- Microsoft is developing its own AI models to offer alternatives.
- New models aim to match Anthropic's performance at lower costs.
- Company seeks to reduce reliance on expensive external AI providers.
Microsoft is intensifying efforts to build its own artificial intelligence models, positioning cost-effectiveness as a key differentiator. The company recently unveiled seven new AI models at its annual Build conference, with a stated goal of matching the performance of existing premium models, such as Anthropic's Opus 4.6, but at a significantly lower price point. This strategy directly addresses the escalating AI spending observed across industries, prompting companies like Walmart and Uber to implement spending caps.
Mustafa Suleyman, leading Microsoft's in-house AI model development, highlighted the substantial expense associated with external AI solutions like Anthropic's. The new models are designed to reduce and eventually eliminate these costs for Microsoft, which itself consumes vast numbers of AI tokens for products like Copilot and internal development. This initiative also aims to mitigate risks associated with long-term contracts for AI services, such as the current arrangement with OpenAI, and establish Microsoft's independent AI capabilities.
Beyond new AI models, Microsoft's Build conference also featured advancements in workplace AI software, a quantum chip, and AI-enhanced employee badges. These developments underscore Microsoft's broad AI ambitions and its focus on leveraging AI across its product ecosystem and internal operations to drive efficiency and innovation.
