He Worldwide spending on cloud infrastructure services reached 102.6 billion of dollars in the third quarter of 2025which represents a year-on-year increase of 25%. It is the fifth consecutive quarter in which this aspect grows above 20%.
This sustained growth reflects, according to the consulting firm Omdia, a change in the technology sector, since the demand for AI by companies is moving from early experimentation to its implementation on an industrial scale.
As this shift accelerates, hyperscalars are increasingly reining in the competition, moving it away from incremental gains in model performance toward platform-level capabilities that support multiple model deployments, as well as ensuring real-world AI agents operate reliably.
In the period analyzed, both AWS, Azure and Google Cloud maintained their positions in the market compared to the second quarter of the year, in addition to accounting for 66% of global spending on cloud infrastructure. Added together, the three had a year-on-year growth of 29%.
AWS growth rose to 20% year-on-year between July and September, its best result since 2022. Azure and Google Cloud also rose in revenue by more than 35% in each case. All three recorded further increases in third-quarter backlogs.
On the other hand, hyperscalar AI strategies are changing: from an approach focused on the incremental performance of the models they are moving to more platform-focused and production-ready approaches. Additionally, companies no longer evaluate AI platforms solely on model capabilities, and increasingly do so on their support for multi-model strategies and agent-based applications.
This is leading hyperscalars to platform-level AI capabilities, and the top three are integrating their own core models with more and more open-weight and third-party models. Additionally, they are increasing investment in agent creation and execution capabilities.
AWS held first place in cloud infrastructure during the third quarter of 2025, with a market share of 32% and year-on-year revenue growth of 20%. As for Azure, it remained in second position, with a market share of 22% and revenue growth of 40%. As for Google Cloud, it reported year-over-year revenue growth of 36% and improved its market share to 11%, primarily due to its enterprise AI offerings.
Rachel Brindley, Senior Director at Omdiarecalled that «Collaboration across the ecosystem remains essential. Multi-model support is increasingly seen as a production requirement rather than a feature as enterprises seek resiliency, cost control, and deployment flexibility in generative AI workloads.«.
