By Vignesh R
Publication Date: 2026-01-30 13:56:00
Currently, many AI startups are under pressure to secure reliable cloud capacity as demand for powerful models continues to rise. That race has pushed companies to sign large, long-term infrastructure deals to make sure they can scale without disruption.
Recently, Perplexity has signed a $750 million cloud agreement with Microsoft, according to sources. The agreement runs for three years and will see Perplexity use Microsoft’s Azure cloud services. Under the deal, the NVIDIA-backed startup will be able to run a range of AI models through Microsoft’s Foundry platform, including models from OpenAI, Anthropic, and xAI.
Multi-cloud strategy stays intact
A Perplexity spokesperson told Bloomberg that the company is partnering with Microsoft to gain access to “frontier models from X, OpenAI and Anthropic.” The spokesperson also said the deal does not replace Perplexity’s existing cloud setup.
Amazon Web Services remains Perplexity’s main cloud provider, and the company has not shifted spending away from AWS as part of the Microsoft agreement.
The deal highlights how fast-growing AI companies are increasingly adopting multi-cloud strategies rather than relying on a single provider. By spreading workloads across platforms, startups can improve reliability and gain access to a wider range of AI models.
For Microsoft, the agreement adds another high-profile AI customer to Azure as it positions Foundry as a central hub for deploying models from…