Artificial intelligence has dramatically accelerated how developers write code—but faster coding alone has not translated into proportionally faster innovation. GitLab’s newly announced GitLab Duo Agent Platform, now generally available, directly addresses this challenge by introducing agentic AI orchestration across the entire DevSecOps lifecycle, not just the coding phase.
Designed for modern software teams, the platform helps organizations overcome what many now call the AI paradox in software delivery: when AI boosts code creation speed, but downstream bottlenecks slow everything else.
Solving the AI Paradox in Software Delivery
While developers report productivity gains of up to 10x using AI-assisted coding tools, coding represents only about 20% of the software development lifecycle. The remaining effort—code reviews, security validation, compliance checks, CI/CD troubleshooting, and bug remediation—often becomes more congested as code velocity increases.
GitLab Duo Agent Platform tackles this imbalance by enabling intelligent, policy-aware AI agents that operate across planning, development, security, and deployment workflows. Instead of isolated AI tools, teams gain coordinated agents that work within GitLab’s system of record, aligned with organizational standards and governance controls.
Agentic Chat: Context-Aware AI Where Developers Work
At the core of the platform is Agentic Chat, a context-aware assistant embedded across the GitLab Web UI and popular IDEs. Using multi-step reasoning, it answers complex questions and performs autonomous actions informed by real project data—including issues, merge requests, pipelines, and security findings.
Key use cases include:
Analyze and Understand
Agentic Chat can summarize issues and merge requests, highlight key risks, and help developers understand unfamiliar codebases, architectures, and dependencies—directly within repositories or IDEs.
Code and Modernize
From generating application code and infrastructure-as-code to fixing bugs, modernizing legacy systems, and producing documentation, Agentic Chat supports a wide range of languages and frameworks. It integrates seamlessly with VS Code, JetBrains IDEs, Cursor, and Windsurf, with customizable rules at the user or workspace level.
CI/CD and Security Support
The assistant simplifies pipeline creation and troubleshooting, explains vulnerabilities in plain language, prioritizes security findings, and recommends actionable fixes—helping teams maintain velocity without compromising security.
Foundational and Custom AI Agents Built for DevSecOps
The general availability release introduces Foundational Agents, prebuilt by GitLab experts to manage complex lifecycle tasks:
- Planner Agent streamlines work breakdown, prioritization, and execution.
- Security Analyst Agent reviews vulnerabilities, clarifies impact, and guides remediation.
For greater flexibility, teams can build Custom Agents using GitLab’s AI catalog, tailoring context, capabilities, and workflows to match internal engineering standards and guardrails.
GitLab also supports External Agents, including Claude Code from Anthropic and Codex CLI from OpenAI, delivering native access with transparent security controls and embedded LLM subscriptions.
Agentic Flows That Automate End-to-End Work
Foundational Agentic Flows chain multiple agents to automate complex processes, such as converting issues into merge requests, modernizing CI/CD pipelines, fixing failed pipelines, and accelerating AI-assisted code reviews. These flows enable consistent, repeatable execution across development teams.
Governance, Deployment Flexibility, and Transparent Usage
Available on GitLab.com, GitLab Self-Managed, and GitLab Dedicated (with full availability during the GitLab 18.8 release cycle), the platform offers robust governance features. Administrators gain visibility into agent usage, actions, and impact, ensuring responsible AI adoption at scale.
Organizations can select AI models at the namespace level, support self-hosted models in self-managed deployments, and control access via group-based policies with LDAP and SAML integration.
Pricing and Getting Started
GitLab Duo Agent Platform uses GitLab Credits, a new usage-based pricing model. Premium and Ultimate customers receive monthly included credits—$12 and $24 per user, respectively—at no additional cost, with flexible options to purchase more as needed.
The Future of Agentic AI in Software Delivery
As IDC forecasts that 70% of organizations will embed AI agents into DevOps and DevSecOps pipelines by 2030, GitLab Duo Agent Platform positions itself as a unified orchestration layer for scalable, governed AI-driven software delivery.
About GitLab
GitLab is the most comprehensive, intelligent DevSecOps platform for software innovation. GitLab enables organizations to increase developer productivity, improve operational efficiency, reduce security and compliance risk, and accelerate digital transformation. More than 50 million registered users and more than 50% of the Fortune 100** trust GitLab to ship better, more secure software faster.
