Skip to content
InfoQ·

🤖GitLab Duo Self-Hosted Supports Microsoft Foundry Models

Self-hosted AI gets a Microsoft boost

TL;DR

GitLab Duo Self-Hosted now integrates with Microsoft Foundry, enabling support for models from OpenAI, Anthropic, Meta, and Mistral. This move is crucial for organizations needing data residency and regulatory compliance.

GitLab Duo Self-Hosted has expanded to support models deployed through Microsoft Foundry, including OpenAI GPT, Anthropic Claude, Meta Llama, and Mistral. This integration is particularly relevant for organizations with strict data residency, sovereignty, and regulatory requirements. The architecture includes a self-managed GitLab instance, a self-hosted AI Gateway, and model endpoints hosted through Microsoft Foundry. This setup allows organizations to use their own AI Gateway and model deployments, providing greater flexibility but also more responsibility for engineering teams. The key trade-off is that while organizations gain control over models and infrastructure, they must verify compatibility across both platforms before selecting a model.

GitLab Duo Self-Hosted Supports Microsoft Foundry Models — InfoQ

Key Points

1

GitLab Duo Self-Hosted supports models from OpenAI, Anthropic, Meta, and Mistral through Microsoft Foundry.

2

The integration is crucial for organizations with data residency, sovereignty, and regulatory requirements.

3

The architecture includes a self-managed GitLab instance, a self-hosted AI Gateway, and model endpoints from Microsoft Foundry.

4

Organizations can use different models for different GitLab Duo capabilities without changing the development workflow.

5

Microsoft Foundry provides access to models from multiple vendors, giving enterprises more choice and control.

Why It Matters

If you're an enterprise with strict data residency and regulatory requirements, GitLab's new Microsoft Foundry integration is a game-changer. It allows you to use your own AI Gateway and model deployments, providing greater flexibility but also more responsibility for engineering teams. The key is to verify compatibility across both platforms before selecting a model.

gitlabmicrosoft foundryai modelsdata residencyregulatory compliance

Comments

Subscribe to join the conversation...

Be the first to comment

Enjoyed this article?

Get it daily. 7am. Free. Reads in 5 minutes.

Join 3,567 builders reading daily.