Optimize your team's price-performance with hosted open weight models

To search repositories and projects, login to gitlab.com . Suggestions GitLab Duo Agent Platform Code Suggestions (AI) CI/CD GitLab on AWS GitLab on Google Cloud Why GitLab?

What happened

To search repositories and projects, login to gitlab. com . Platform Execution & Workflows CI/CD Source Code Management Agile Delivery Security & Governance Application Security Testing Governance & Compliance Supply Chain Security Context & AI Agentic Orchestration Context Graph Visibility & Measurement Explore the Platform Why GitLab One platform for speed with control across your software lifecycle.

Where the sources line up

Solutions Outcomes DevOps Modernization Security Modernization AI Modernization By size Enterprise Small Business Startups Industries Financial Services Public Sector Telecommunications Automotive Education Aerospace View all Solutions GitLab Transcend Catch our latest innovations announced at the last Transcend.

Practical impact for readers

Three open weight models are now available and hosted on GitLab Duo Agent Platform, so you can tune quality, latency, and cost per workload, within your guardrails.

Who should pay attention now

There’s no single best model for every software development task. Implementing a new feature, diagnosing a failed pipeline, and resolving security vulnerabilities all place different demands on the model handling them. GitLab Duo Agent Platform is expanding GitLab-managed model choice with three hosted open weight models: Kimi K3, GLM 5. 3, and MiniMax M3.

What is still unclear

Together with the frontier models already available in GitLab, your team now has more control over how you optimize for quality, latency, and cost, tuned to the needs of each workload. Point a hard task at Kimi K3 or GLM 5. 3, which outperformed comparable frontier models in internal testing and costs less per call, or hand routine, high-volume work to MiniMax M3. Either way, your team gets up to 4x more calls per GitLab Credit than some comparable frontier models, with more AI model options to match cost to task complexity.

Source notes

GitLab Blog