GitLab Dedicated AI Gateway Data Residency Guide
On this page
Quick answer
GitLab announced Dedicated AI Gateway generally available for GitLab Dedicated, describing agentic workloads inside the same single-tenant environment and region. Validate that deployment-specific contract for the tenant rather than relying on generic AI Gateway wording.
GitLab’s general AI Gateway documentation says multi-region routing is not strict data sovereignty: traffic can move for latency or availability and requests are not guaranteed to remain in one region. The Dedicated path and generic cloud path must not be conflated.
Residency evidence
Record the GitLab deployment type, Gateway placement, inference provider and model, prompts and outputs, repositories and artifacts, logs, backups, support access, subprocessors, encryption, routing and outage behavior, telemetry, retention, deletion, and contractual commitments.
Test an unsupported model, regional disruption, provider fallback, support investigation, logging export, user revocation, and deletion. A single-tenant application region does not automatically prove every model call, log, or support path stays there.
Continue with the 19.3 overview and readiness checklist.
Official sources
Source check: August 23, 2026. Verify the tenant’s Dedicated architecture, models, Regions, routing, logs, support, and contract.