Skip to content

Models Registry

The Models Registry is where users can connect and manage LLMs and embedding models from different providers. These models power workflows, agents, and RAG pipelines throughout the platform. You can register new models with API credentials, assign them to agents or bots, and monitor their connectivity status.

Key Features

  • LLM and Embedding Model Support: Register both large language models and embedding models for use across workflows and RAG bots.
  • Multi-Provider Integration: Supports models from various providers by allowing flexible input for model name, API key, host, and model ID.
  • Connectivity Checks: Verify model reachability before saving.
  • Model Permissions: Control access to specific models for teams or projects.
  • Status Monitoring: See real-time connection status (Connected / Disconnected) to quickly debug issues.