GitLab + Google Vertex integration

Software teams using GitLab can now connect its artificial intelligence agents to Google Cloud’s Vertex AI models, including Gemini, without leaving GitLab’s control layer. (about.gitlab.com) GitLab announced the expanded Google Cloud partnership on April 14, 2026. The companies said GitLab Duo Agent Platform can now call Vertex AI natively, and Google Cloud customers can count that usage against existing cloud commitments. (about.gitlab.com / ir.gitlab.com) Vertex AI is Google Cloud’s service for accessing large language models and other machine learning tools. GitLab said its agents can use those models while actions still pass through GitLab permissions, approvals, compliance policies, and audit trails. (cloud.google.com / about.gitlab.com) GitLab’s pitch is that coding agents are more useful when they can see the full software workflow, not just a text prompt. The company said GitLab remains the system of record for issues, code, merge requests, pipelines, and security findings, so an agent can work from that context inside one platform. (about.gitlab.com / about.gitlab.com) The new arrangement also changes the infrastructure piece. GitLab said organizations can run its artificial intelligence gateway on Google Cloud managed or self-hosted runtimes, including Google Kubernetes Engine and Cloud Run, instead of standing up separate model-serving infrastructure. (about.gitlab.com / tech.yahoo.com) That matters for companies trying to deploy coding agents under internal rules on spending, data handling, and auditability. GitLab said customers can choose models from Vertex AI Model Garden based on performance, cost, or regulatory requirements while keeping governance in GitLab. (about.gitlab.com / cloud.google.com) GitLab has been building toward this for months. It introduced GitLab Duo Agent Platform in June 2025, announced general availability on January 15, 2026, and expanded access in March through a credit-based pricing model that also opened the platform to GitLab.com free-tier organizations. (about.gitlab.com / about.gitlab.com / about.gitlab.com) Google Cloud has been pushing Vertex AI as the layer where enterprises pick among first-party and third-party foundation models. GitLab’s update slots its own agent software into that stack, with GitLab handling workflow and controls while Google Cloud supplies the model access and compute. (cloud.google.com / about.gitlab.com) The immediate question is whether large customers want one vendor for code workflow and another for models, or a single stack from a cloud provider. GitLab and Google are betting that enterprises will pay for that split if it keeps artificial intelligence agents inside existing approval and audit systems. (about.gitlab.com / cloud.google.com)