AI providers

Understand the Model Provider Registry

Manage provider definitions, profiles, capabilities, credentials, local endpoints, and default models from one registry.

9 min readIntermediateUpdated 2026-07-29For LobeWork 1.0+
01

Definitions and profiles are separate

A provider definition describes capabilities and transport. A profile stores the user-specific configuration for one account or endpoint.

02

Credentials stay protected

Provider secrets should use the secure credential boundary rather than renderer storage or ordinary project files.

03

Capabilities guide the UI

Chat, streaming, vision, embeddings, image generation, tool calls, local use, and model discovery can vary by provider.

04

Defaults remain visible

Default models should be chosen deliberately and confirmed when a voice command or workflow proposes a change.

05

Extensions can contribute metadata

Reviewed official Extension manifests can contribute provider definitions without enabling arbitrary third-party code.

NEED A GUIDED ANSWER?

Ask LobeWork about this guide.

Ask LobeWork uses the same published article library and links back to its source guidance.

WAS THIS GUIDE HELPFUL?Help us improve the documentation.