Crate elevenlabs_rs

Expand description

An unofficial lib crate for ElevenLabs.

§Example

§Text to Speech

use elevenlabs_rs::{ElevenLabsClient, Result, DefaultVoice, Model};
use elevenlabs_rs::endpoints::genai::tts::{TextToSpeech, TextToSpeechBody};
use elevenlabs_rs::utils::play;

#[tokio::main]
async fn main() -> Result<()> {
    let client = ElevenLabsClient::from_env()?;

    let txt = "Hello! 你好! Hola! नमस्ते! Bonjour! \
        こんにちは! مرحبا! 안녕하세요! Ciao! Cześć! Привіт! வணக்கம்!";

    let body = TextToSpeechBody::new(txt)
       .with_model_id(Model::ElevenMultilingualV2);

    let endpoint = TextToSpeech::new(DefaultVoice::Brian, body);

    let speech = client.hit(endpoint).await?;

    play(speech)?;

    Ok(())
}

§WebSocket lifecycle

WebSocket entry points such as ElevenLabsClient::connect_text_to_speech, ElevenLabsClient::connect_multi_context_text_to_speech, and ElevenLabsClient::connect_realtime_speech_to_text return a WebSocketSession when the ws and genai features are enabled. The session implements Stream for inbound server messages and owns background reader/writer tasks.

Prefer calling WebSocketSession::close when the session has completed so the client sends a close frame before shutting down background work. Dropping the session is still safe and aborts background tasks as a fallback. For diagnostics after a stream ends, or after close/abort, call WebSocketSession::join to inspect reader and writer task completion. Use WebSocketOptions with the connect_*_with_options methods to tune inbound buffering and the graceful-close timeout.

Modules§

endpoints
error
multipart: multipart/form-data
utils

Macros§

pin_mut: Pins a value on the stack.

Structs§

ApiResponse
Bytes: A cheaply cloneable and sliceable chunk of contiguous memory.
ElevenLabsClient
Method: The Request Method (VERB)
RawRequestBuilder
VoiceSettings: Voice settings
WebSocketOptions: Runtime settings for ElevenLabs WebSocket sessions.
WebSocketSession: A live ElevenLabs WebSocket session.
WebSocketSessionReport: Completion report for the reader and writer tasks owned by a session.

Enums§

DefaultVoice
FilePart: Source for a multipart file field.
Language: Language
LegacyVoice
Model
OutputFormat: See Elevenlabs documentation on supported output formats.
WebSocketTaskStatus: Completion status for one background WebSocket task.

Traits§

StreamExt: An extension trait for Streams that provides a variety of convenient combinator functions.

Type Aliases§

Result