GoogleVertexProvider

Namespace: Vercel.AI.Google

public sealed class GoogleVertexProvider : GoogleProvider, IBatchProvider

Google Vertex AI provider.

Inheritance: object → ProviderBase → GoogleProvider

Implements: IBatchProvider

View source

Constructors

GoogleVertexProvider(HttpClient, VertexOptions?)

public GoogleVertexProvider(HttpClient httpClient, VertexOptions? options = null)

Creates a Vertex provider.

Parameters

Name	Type	Description
httpClient	HttpClient
options	VertexOptions

View source

Properties

Vertex

public VertexOptions Vertex { get; }

Vertex options.

Value: VertexOptions

View source

Methods

Create(VertexOptions?, HttpMessageHandler?)

public static GoogleVertexProvider Create(VertexOptions? options = null, HttpMessageHandler? handler = null)

Creates a provider.

Parameters

Name	Type	Description
options	VertexOptions
handler	HttpMessageHandler

Returns: GoogleVertexProvider

View source

Interactions(string)

public override ILanguageModel Interactions(string modelId)

Interactions use the location resource and reject Express Mode.

Parameters

Name	Type	Description
modelId	string

Returns: ILanguageModel

View source

Interactions(GoogleInteractionsAgent)

public override ILanguageModel Interactions(GoogleInteractionsAgent agent)

Interactions use the location resource and reject Express Mode.

Parameters

Name	Type	Description
agent	GoogleInteractionsAgent

Returns: ILanguageModel

View source

LanguageModel(string)

public override ILanguageModel LanguageModel(string modelId)

Tuned endpoints/ models reject Express Mode and omit /publishers/google.

Parameters

Name	Type	Description
modelId	string

Returns: ILanguageModel

View source

SpeechModel(string)

public override ISpeechModel SpeechModel(string modelId)

Chirp speech models use Cloud Text-to-Speech. Gemini speech models use generateContent.

Parameters

Name	Type	Description
modelId	string

Returns: ISpeechModel

View source

TranscriptionModel(string)

public override ITranscriptionModel TranscriptionModel(string modelId)

Gemini transcription ids use generateContent. Other ids use Cloud Speech-to-Text.

Parameters

Name	Type	Description
modelId	string

Returns: ITranscriptionModel

View source