Provider Architecture

The AI SDK uses a layered provider architecture that follows the adapter pattern, enabling support for multiple AI providers through a unified interface.

Architecture Overview

mermaid

graph LR
    AI["Main Package
ai"] --> PROVIDER["Specifications
@ai-sdk/provider"]
    AI --> UTILS["Shared Utilities
@ai-sdk/provider-utils"]
    UTILS --> PROVIDER
    OPENAI["Provider
e.g. @ai-sdk/openai"] --> PROVIDER
    OPENAI --> UTILS
    OPENAI --> OPENAI_API["Provider API
e.g. OpenAI API
(External)"]

    classDef core fill:#f3e5f5,stroke:#4a148c,stroke-width:2px
    classDef spec fill:#fff3e0,stroke:#e65100,stroke-width:2px
    classDef utils fill:#e8f5e8,stroke:#1b5e20,stroke-width:2px
    classDef provider fill:#fff8e1,stroke:#ff6f00,stroke-width:2px
    classDef external fill:#f1f8e9,stroke:#33691e,stroke-width:2px

    class AI core
    class PROVIDER spec
    class UTILS utils
    class OPENAI provider
    class OPENAI_API external

See the provider abstraction architecture doc for a more in-depth guide.