Skip to main content

README

GitLab Duo - AI-Powered DevSecOps

Complete documentation for GitLab Duo and AI features in GitLab Ultimate.

Overview

GitLab Duo is GitLab's comprehensive AI-powered platform that accelerates software development across the entire DevSecOps lifecycle. Starting with GitLab 18.0, Code Suggestions and Chat are included with Premium and Ultimate tiers at no additional cost.

GitLab Duo Agent Platform GA: January 2026 (GitLab 18.8) Global Launch Event: February 10, 2026

Documentation Structure

[object Object]

Complete overview of GitLab Duo features, capabilities, and availability.

Topics:

  • What is GitLab Duo
  • Key components (Code Suggestions, Chat, Agent Platform)
  • Tier availability (Premium vs Ultimate)
  • Model selection
  • Self-hosted deployment
  • Getting started

Read this first to understand the full scope of GitLab Duo.

[object Object]

Deep dive into AI-powered code generation and assistance.

Topics:

  • Code Suggestions (real-time completion)
  • GitLab Duo Chat (conversational AI)
  • Supported languages (14+)
  • IDE integration (VS Code, JetBrains, etc.)
  • Test generation
  • Code explanation and documentation
  • Refactoring assistance

Use cases:

  • Writing new code faster
  • Understanding unfamiliar codebases
  • Generating tests automatically
  • Improving code quality

[object Object]

AI-powered pipeline automation and optimization.

Topics:

  • Root Cause Analysis (identify pipeline failures)
  • Pipeline Fix Suggestions (automatic remediation)
  • Intelligent Test Selection
  • Flaky Test Detection
  • Pipeline optimization recommendations
  • Auto DevOps with AI
  • Convert CI File Flow (migrate from other platforms)

Impact: 30% faster releases with AI-powered CI/CD

[object Object]

AI-enhanced security scanning and vulnerability management.

Topics:

  • SAST False Positive Detection (GitLab 18.7+)
  • Vulnerability Explanations
  • Remediation Suggestions
  • Secret Detection with validation
  • Container Scanning
  • License Compliance
  • AI Code Review for security

Benefits:

  • Faster vulnerability triage
  • Reduced false positives
  • Automated remediation
  • Better security coverage

[object Object]

AI-assisted project planning and issue management.

Topics:

  • Issue Description Generation
  • Planner Agent (Beta - create and edit)
  • Merge Request Summaries
  • Discussion Summaries
  • Work Breakdown Suggestions
  • Sprint Planning Assistance
  • Roadmap Planning

Time savings:

  • Issue writing: 60% faster
  • Sprint planning: 40% faster
  • Better quality requirements

[object Object]

GitLab Duo Agent Platform - build and orchestrate AI agents.

Topics:

  • Foundational Agents (Developer, Security, Planner, Tester)
  • Custom Agents (build your own)
  • External Agents (Claude, GPT, Mistral)
  • Flows (multi-agent workflows)
  • Pre-built Flows (Software Development, Issue-to-MR, Convert CI)
  • Model Context Protocol (MCP) integration
  • AI Catalog (share agents)

Platform GA: January 2026 (GitLab 18.8)

[object Object]

Programmatic access to GitLab Duo features.

Topics:

  • Authentication (PAT, OAuth2, CI/CD tokens)
  • Code Suggestions API
  • Chat API
  • Security API (vulnerability explanations)
  • CI/CD AI API (root cause analysis)
  • Agent Platform API (execute flows)
  • Model Selection API
  • Webhooks

Use cases:

  • Custom IDE integrations
  • Automation scripts
  • Third-party tool integration
  • Workflow extensions

[object Object]

When and how to use AI effectively.

Topics:

  • When to use AI (excellent use cases)
  • When NOT to use AI (critical code)
  • Prompt engineering techniques
  • Security best practices
  • Performance considerations
  • Cost management
  • Privacy and compliance
  • Team adoption strategies
  • Testing AI-generated code

Essential reading for effective AI usage.

Quick Start

1. Enable GitLab Duo

For Premium/Ultimate users:

Settings  GitLab Duo  Enable features

For Duo Pro users ($19/user/month):

Subscribe  Enable in group settings

2. IDE Setup

VS Code:

  1. Install GitLab Workflow extension
  2. Sign in with GitLab account
  3. Enable Code Suggestions
  4. Start coding

JetBrains:

  1. Install GitLab plugin
  2. Configure authentication
  3. Enable AI assistance
  4. Use right-click menu

3. Try Core Features

Code Suggestions:

# Start typing def calculate_ # AI suggests complete function def calculate_total(items: List[Item]) -> float: return sum(item.price for item in items)

Chat:

You: Explain how JWT authentication works

Duo: JWT (JSON Web Token) authentication works by...

Root Cause Analysis:

Pipeline fails  Click job  View Root Cause Analysis

Key Features by GitLab Version

GitLab 18.0 (May 2025)

  • Code Suggestions included in Premium/Ultimate
  • Duo Chat included in Premium/Ultimate
  • No additional subscription required

GitLab 18.1 (June 2025)

  • Codestral model for Code Suggestions
  • MR Summary model updated to Claude 4.0 Sonnet

GitLab 18.2 (July 2025)

  • Claude Sonnet 4 as default for Code Suggestions
  • Improved performance and accuracy

GitLab 18.4 (September 2025)

  • Model Selection feature (Beta)
  • GitLab Knowledge Graph

GitLab 18.5 (October 2025)

  • Model Selection GA
  • Pipeline Fix Suggestions default enabled

GitLab 18.7 (December 2025)

  • SAST False Positive Detection
  • Planner Agent Beta (create and edit)
  • Duo Analytics dashboard (6-month trends)
  • Secret validity checks

GitLab 18.8 (January 2026)

  • GitLab Duo Agent Platform GA
  • Full multi-agent orchestration
  • Custom flows and agents
  • February 10, 2026: Global launch event

Adoption Statistics (2025)

  • 1.5 million developers using GitLab AI tools
  • 30% faster releases with AI-powered CI/CD
  • 60% faster issue writing with AI generation
  • 70% faster test creation with AI
  • 88% success rate for Software Development Flow

Use Case Examples

Feature Development

1. Create issue with AI-generated description
2. Use Software Development Flow
3. AI generates code + tests
4. Security scan + code review by AI
5. Human reviews MR
6. Deploy

Time savings: 40-70% depending on complexity

Bug Investigation

1. Pipeline fails
2. Root Cause Analysis identifies issue
3. AI suggests fix
4. Auto-generates fix MR
5. Tests verify fix
6. Deploy

Time savings: 50% for investigation + fix

Security Review

1. SAST scan runs
2. AI identifies false positives (90% confidence)
3. Team focuses on real vulnerabilities
4. AI suggests remediation
5. Verify and apply

Time savings: 25% in triage + remediation

Cost Considerations

Included with Subscription

Premium/Ultimate (GitLab 18.0+):

  • Code Suggestions: Unlimited
  • Duo Chat: Unlimited
  • Agent Platform (Ultimate only)
  • No per-request charges

Standalone Duo Pro

  • $19/user/month
  • All Duo features
  • For organizations without Premium/Ultimate

Self-Hosted Costs

  • Infrastructure (compute, storage)
  • Model hosting (if cloud-based)
  • Network bandwidth
  • Maintenance

Typical savings: 10-20% with optimization

Security and Privacy

Data Handling

Sent to AI:

  • Code context (limited)
  • File structure
  • Issue/MR descriptions

NOT sent:

  • Full repository
  • Secrets (if detected)
  • Other projects
  • Personal data

Self-Hosted Option

For regulated industries:

  • Complete data control
  • No external API calls
  • Air-gapped deployments
  • Full audit logs

Industries: Healthcare, finance, government, automotive

Compliance

  • GDPR: Data retention controls
  • HIPAA: Self-hosted deployment
  • SOC 2: Audit logging and access controls
  • PCI-DSS: Secure development practices

Getting Help

Documentation

Support

  • GitLab Support Portal
  • Community Forum
  • Issue Tracker

Training

  • GitLab Learn
  • Professional Services
  • Team Workshops

Contributing

Found an issue or have suggestions for this documentation?

  1. Create issue in GitLab
  2. Tag: documentation, gitlab-duo
  3. Assign to: Documentation team

Version History

VersionDateChanges
1.02026-01-08Initial comprehensive documentation

Last Updated: 2026-01-08 Next Review: 2026-04-08


Ready to get started? Begin with Overview to understand all available features.

Need practical guidance? Jump to Best Practices for when and how to use AI effectively.

Want to build custom agents? Check out Agents for the Agent Platform guide.