Google Vertex AI

Vertex AI is Google Cloud's platform for training, tuning, evaluating and serving models, including the Gemini family, partner models from Anthropic and Mistral, embeddings, vector search and agent tooling, billed per token, per character or per compute hour. It has run in the Dammam region since May 2024, with Assured Workloads sovereign controls operated by the Saudi reseller CNTXT together with SITE, and Google's locations page lists Dammam and Doha under Middle East with a shared list of Gemini models. The data-residency page, updated 7 November 2026, is the document to read: data stored at rest stays in the chosen region, but Google's in-region machine-learning processing commitment covers thirteen locations in the Americas, Europe and Asia and does not name Dammam or Doha. Global endpoints carry no residency guarantee at all. Gemini Enterprise, the agent product, lists no Middle East data-residency region.