GitLab runs agents on Google

GitLab announced enterprise AI coding agents can run directly on Google models without customers hosting separate model infrastructure. The integration reduces infra friction for agentic DevSecOps workflows while tying agent compute to Google Cloud model hosting. (stocktitan.net)

GitLab said on April 14 that its coding and security agents can now run on Google Cloud’s Vertex AI models, including Gemini, without customers standing up separate model infrastructure. (about.gitlab.com) The change plugs GitLab Duo Agent Platform into Vertex AI, Google Cloud’s service for calling large language models, so GitLab agents can use models customers already run in Google Cloud. GitLab said organizations can also run its AI Gateway on Google Cloud instead of provisioning a separate artificial intelligence stack. (about.gitlab.com) GitLab said agent actions will still be governed by its existing compliance and audit controls, and Google Cloud customers can count Duo Agent Platform usage against their committed cloud spend. Google Cloud said the setup gives customers access to Vertex AI’s model catalog and enterprise controls inside GitLab workflows. (about.gitlab.com) GitLab’s agents are software helpers that can plan work, write or review code, and run security or analytics tasks across the software pipeline. GitLab put Duo Agent Platform into general availability on January 15, 2026, with prebuilt agents for planning, coding, security analysis, analytics, and custom workflows triggered by GitLab events. (about.gitlab.com) The company has been pushing artificial intelligence beyond chat-style code suggestions and into “agentic” workflows that take actions inside source control, continuous integration, and security review. GitLab says that model choice, permissions, and traceability are central for large companies that want those agents to operate inside regulated development processes. (about.gitlab.com) The Google tie-up extends a partnership that started in May 2023, when GitLab said it would use Google Cloud generative artificial intelligence for features such as vulnerability explanation while keeping customer code and intellectual property inside GitLab’s infrastructure. The new announcement shifts more of the model execution layer directly onto Google Cloud services. (about.gitlab.com) Investors treated the news as material. MarketWatch reported GitLab shares were up about 5% to $21.10 in after-hours trading on April 14 after the company announced the expanded collaboration. (marketwatch.com) The practical effect is that companies already standardized on Google Cloud can buy GitLab’s agents without also operating a separate model-serving layer. That leaves GitLab selling the workflow and governance layer while Google Cloud supplies the underlying model hosting. (about.gitlab.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.