Large enterprises replace expensive external AI APIs with in-house or open-source models to cut costs and keep data in house. Microsoft leads the trend by substituting OpenAI/Anthropic with its own MAI models.
Redmond's giant moves its apps from external models to in-house ones. A move that redefines API dependency and signals a shift toward direct inference control, with ripple effects on costs, latency, and deployment choices.
Microsoft is the latest member of big tech to trim AI spending by turning to in-house models. Beyond the immediate savings, this move signals a maturing ecosystem where smaller, self-trained models can fulfill enterprise needs, opening the door to on...
According to Bloomberg, Microsoft has started using its own MAI models instead of OpenAI and Anthropic in some product features, driven by cost and data residency considerations. The shift is incremental, with most Copilot traffic still handled by ex...
With API prices for OpenAI and Anthropic climbing, US enterprises are increasingly evaluating Chinese open models. The trend reshapes market dynamics, forcing a rethink of deployment architectures: self-hosting is returning to the fore, balancing TCO...
