Microsoft is moving away from OpenAI models to reduce costs

The US corporation Microsoft has begun phasing out external models provided by OpenAI and Anthropic to optimize the massive costs associated with AI development. According to Bloomberg, the tech giant is now relying on its own proprietary solutions. This was reported by Ixbt.com reports.

Currently, a number of user requests in popular office applications like Excel and Word are being processed using Microsoft's internal family of AI models. Previously, the company emphasized that most features in Microsoft 365 services were powered by OpenAI technologies, but the issue of economic efficiency has led to a change in strategy.

Microsoft does not plan to abandon third-party solutions entirely, but it is implementing its own neural networks wherever cost-cutting opportunities exist. In recent months, the company has significantly increased its activity in this area. Specifically, at the Build conference, 7 new AI models specialized in programming and text-to-image generation were unveiled.

A new trend in the AI market

Experts believe that processing Large Language Models (LLM) remains one of the most expensive operations for technology companies today. Although Microsoft has not officially commented on these changes, analysts view this as an attempt to reduce the cost of cloud AI services.

Microsoft is not alone in this endeavor. According to reports, the following companies are also reviewing their AI expenses:

• Amazon — striving for independence by developing its own internal chips and models;
• Meta — saving on licensing fees by using open-source Llama models;
• Uber and Accenture — seeking alternative solutions to reduce operational costs.

High prices in the AI market are even forcing some Western companies to consider using cheaper Chinese models, despite security concerns. For users in Uzbekistan, these changes may go unnoticed, but in the long term, they will help make Microsoft services more stable and affordable.

In conclusion, while the partnership between Microsoft and OpenAI remains, the push for technological independence is intensifying. This indicates that a race for not only quality but also economic efficiency has begun in the field of artificial intelligence.