Class AudioCppClient

Namespace

Mabs.AudioCpp

Assembly

AudioCpp.dll

Client for the audio.cpp server protocol. One instance maps to one server base URL and is thread-safe; share it. The per-area sub-clients (Speech, Transcriptions, Alignments, Tasks, Recipes) carry the endpoint groups; readiness, model, voice and unload routes sit directly on this type.

public sealed class AudioCppClient

Inheritance

object

AudioCppClient

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.ReferenceEquals(object, object)

object.ToString()

Constructors

AudioCppClient(string, HttpClient?, string?)

Creates a client from a base URL string.

public AudioCppClient(string baseUrl, HttpClient? httpClient = null, string? apiKey = null)

Parameters

baseUrl string

Server root, for example http://127.0.0.1:8080.

httpClient HttpClient

Optional transport; see the Uri constructor overload.

apiKey string

Optional bearer token sent on every request.

AudioCppClient(Uri, HttpClient?, string?)

Creates a client for the server at baseUrl.

public AudioCppClient(Uri baseUrl, HttpClient? httpClient = null, string? apiKey = null)

Parameters

baseUrl Uri

Server root, for example http://127.0.0.1:8080.

httpClient HttpClient

Optional transport. When omitted, a default HttpClient is created; pass your own to control proxies, timeouts or TLS. For the live routes it must be backed by a handler with full-duplex HTTP/1.1 support (the default handler on .NET Core and .NET Framework's WinHttpClient qualify; browser/WASM handlers do not).

apiKey string

Optional bearer token sent as Authorization: Bearer <key> on every request. The audio.cpp server ignores it today; frontends in front of it may require it.

Properties

Alignments

Forced alignment of known transcripts.

public AlignmentClient Alignments { get; }

Property Value

AlignmentClient

ApiKey

Bearer token sent on every request, when set.

public string? ApiKey { get; }

Property Value

string

Recipes

Multi-model recipes.

public RecipeClient Recipes { get; }

Property Value

RecipeClient

Speech

Speech synthesis, streaming speech, and live speech-to-speech.

public SpeechClient Speech { get; }

Property Value

SpeechClient

Tasks

Generic framework tasks: run, stream, batch.

public TaskClient Tasks { get; }

Property Value

TaskClient

Transcriptions

Transcription, details, streaming, live capture, and native batches.

public TranscriptionClient Transcriptions { get; }

Property Value

TranscriptionClient

Methods

GetHealthAsync(CancellationToken)

Server readiness from GET /health.

public Task<AudioCppHealth> GetHealthAsync(CancellationToken cancellationToken = default)

Parameters

cancellationToken CancellationToken

Cancellation token.

Returns

Task<AudioCppHealth>

The health report.

Exceptions

AudioCppException

The request failed.

ListModelsAsync(bool, bool, CancellationToken)

Configured models from GET /v1/models.

public Task<IReadOnlyList<AudioCppModel>> ListModelsAsync(bool includeParams = false, bool includeSessionOptions = false, CancellationToken cancellationToken = default)

Parameters

includeParams bool

Add each model's declared request options (include_params=true).

includeSessionOptions bool

Echo each model's configured session options (include_session_options=true).

cancellationToken CancellationToken

Cancellation token.

Returns

Task<IReadOnlyList<AudioCppModel>>

The model list (data array).

Exceptions

AudioCppException

The request failed.

ListRecipesAsync(CancellationToken)

List available recipes from GET /v1/recipes.

public Task<IReadOnlyList<RecipeEntry>> ListRecipesAsync(CancellationToken cancellationToken = default)

Parameters

cancellationToken CancellationToken

Cancellation token.

Returns

Task<IReadOnlyList<RecipeEntry>>

The recipe list (data array).

Exceptions

AudioCppException

The request failed.

ListVoicesAsync(string?, CancellationToken)

Discoverable voice names from GET /v1/audio/voices: cached voice ids, configured preset names, and shared voice-dir WAV basenames. Pass model when several models are configured.

public Task<IReadOnlyList<string>> ListVoicesAsync(string? model = null, CancellationToken cancellationToken = default)

Parameters

model string

Model id to scope the lookup to.

cancellationToken CancellationToken

Cancellation token.

Returns

Task<IReadOnlyList<string>>

The voice names the server reports.

Exceptions

AudioCppException

The request failed.

LoadModelAsync(ModelLoadRequest, CancellationToken)

Dynamically load a model through POST /v1/models/load. Only servers built with native model management and started with --ui-management serve this route (otherwise HTTP 403, error type forbidden).

public Task<ModelLoadResult> LoadModelAsync(ModelLoadRequest request, CancellationToken cancellationToken = default)

Parameters

request ModelLoadRequest

The load request (validated before sending).

cancellationToken CancellationToken

Cancellation token.

Returns

Task<ModelLoadResult>

The load report.

Exceptions

AudioCppException

The request failed.

UnloadAllModelsAsync(CancellationToken)

Unload every currently loaded model.

public Task<UnloadModelsResult> UnloadAllModelsAsync(CancellationToken cancellationToken = default)

Parameters

cancellationToken CancellationToken

Cancellation token.

Returns

Task<UnloadModelsResult>

The unload report.

Exceptions

AudioCppException

The request failed.

UnloadModelsAsync(IEnumerable<string>, CancellationToken)

Unload specific models from memory; unknown ids are reported, not an error.

public Task<UnloadModelsResult> UnloadModelsAsync(IEnumerable<string> modelIds, CancellationToken cancellationToken = default)

Parameters

modelIds IEnumerable<string>

Model ids to unload.

cancellationToken CancellationToken

Cancellation token.

Returns

Task<UnloadModelsResult>

The unload report (unloaded and not-found ids).

Exceptions

AudioCppException

The request failed.