SpeechModelStreamCall

Namespace: Vercel.AI.Operations

public sealed class SpeechModelStreamCall

Arguments for doStream on a speech model. Maps to SpeechModelV4StreamOptions.

Inheritance: object

View source

Constructors

SpeechModelStreamCall(SpeechTextStream, string?, SpeechAudioFormat?, string?, double?, string?, JsonElement, IReadOnlyDictionary<string, string>, bool?, CancellationToken)

public SpeechModelStreamCall(SpeechTextStream text, string? voice, SpeechAudioFormat? outputFormat, string? instructions, double? speed, string? language, JsonElement providerOptions, IReadOnlyDictionary<string, string> headers, bool? includeRawChunks, CancellationToken cancellationToken)

Creates a speech stream call.

Parameters

Name	Type	Description
text	SpeechTextStream
voice	string
outputFormat	SpeechAudioFormat
instructions	string
speed	double?
language	string
providerOptions	JsonElement
headers	IReadOnlyDictionary<string, string>
includeRawChunks	bool?
cancellationToken	CancellationToken

View source

Properties

CancellationToken

public CancellationToken CancellationToken { get; }

Cancels the operation. Also fires when the consumer cancels the result stream.

Value: CancellationToken

View source

Headers

public IReadOnlyDictionary<string, string> Headers { get; }

Headers including the user-agent suffix.

Value: IReadOnlyDictionary<string, string>

View source

IncludeRawChunks

public bool? IncludeRawChunks { get; }

When true, providers include raw provider chunks in the stream.

Value: bool?

View source

Instructions

public string? Instructions { get; }

Speaking instructions.

Value: string

View source

Language

public string? Language { get; }

ISO 639-1 language code or auto.

Value: string

View source

OutputFormat

public SpeechAudioFormat? OutputFormat { get; }

Desired format of the output audio chunks. The provider default applies when null.

Value: SpeechAudioFormat

View source

ProviderOptions

public JsonElement ProviderOptions { get; }

Provider options. Empty object when omitted.

Value: JsonElement

View source

Speed

public double? Speed { get; }

Playback speed.

Value: double?

View source

Text

public SpeechTextStream Text { get; }

Text to convert to speech, as a stream of text parts. The provider owns it once doStream is called and releases it on every terminal path.

Value: SpeechTextStream

View source

Voice

public string? Voice { get; }

Voice to use. Provider-specific.

Value: string

View source