Overview

Availability

This capability is available on the Business Plan for the Wren AI Self-Hosted Version.

Use your own LLM provider with Wren AI. During onboarding (or later in Settings), you can connect one or more providers and route different parts of the AI pipeline to the models you prefer.

Setup modes

Quick setup

Use pre-configured templates for popular provider combinations. Perfect for getting started quickly with production‑ready defaults.

• Pre-configured model selections
• Optimized pipeline assignments
• Guided secret configuration
• Automatic testing & validation

Supported quick-start templates

• OpenAI
• Azure OpenAI
• Google Vertex AI
• AWS Bedrock (recommended: Anthropic Claude models)

Manual configuration

Full control over model selection, configuration, and pipeline assignments. Ideal for custom setups and specific requirements.

• Choose any combination of models
• Custom model support
• Fine-tune all parameters
• Flexible pipeline routing