Crate novel_doubao 

Rust library for Doubao (ByteDance Volcengine) APIs.

This crate provides clients for various Doubao APIs:

• TTS: Text-to-Speech API
• ASR: Automatic Speech Recognition API
• Chat: Chat completion API (OpenAI-compatible)
• Embeddings: Text and multimodal embeddings API
• Images: Image generation API
• Tokenization: Text tokenization API

Creating a client

use novel_doubao::Client;
use novel_doubao::config::DoubaoConfig;

// Create a client with default configuration from environment variables.
let client = Client::new();

// Or create with custom configuration.
let config = DoubaoConfig::new()
    .with_app_id("your-app-id")
    .with_api_key("your-api-key")
    .with_access_token("your-access-token")
    .with_resource_id("seed-tts-2.0");
let client = Client::with_config(config);

Making TTS requests

use novel_doubao::Client;
use novel_doubao::spec::tts::CreateSpeechRequestArgs;

let client = Client::new();

let request = CreateSpeechRequestArgs::default()
    .text("Hello, world!")
    .speaker("zh_female_cancan_mars_bigtts")
    .sample_rate(24000u32)
    .build()?;

let response = client.tts().speech().create(request).await?;

// Save to file
response.save("output.mp3").await?;

println!("Generated {} bytes of audio", response.bytes.len());

Making ASR requests

Flash recognition (fastest, single request)

use novel_doubao::Client;

let client = Client::new();

// Recognize from URL
let result = client
    .asr()
    .recognition()
    .flash_url("https://example.com/audio.wav", "user-id")
    .await?;

println!("Recognized: {}", result.result.text);

Standard recognition (for long audio files)

use novel_doubao::Client;

let client = Client::new();

// Submit and wait for result
let result = client
    .asr()
    .recognition()
    .recognize_url("https://example.com/audio.wav", "user-id")
    .await?;

println!("Recognized: {}", result.result.text);

Streaming recognition (real-time)

use bytes::Bytes;
use novel_doubao::Client;
use novel_doubao::spec::asr::StreamingAsrConfigArgs;

let client = Client::new();

let config = StreamingAsrConfigArgs::default().rate(16000u32).build()?;

let mut session = client.asr().streaming().create_session(config).await?;

// Send audio data
session
    .send_audio(Bytes::from_static(b"audio data..."))
    .await?;

// Receive results
while let Some(result) = session.recv().await {
    println!(
        "Partial: {} (final: {})",
        result.result.text, result.is_final
    );
}

Environment Variables

The client reads these environment variables by default:

• DOUBAO_APP_ID: Application ID
• DOUBAO_API_KEY: API key
• DOUBAO_ACCESS_TOKEN: Access token
• DOUBAO_RESOURCE_ID: Resource ID (default: “seed-tts-2.0”)
• DOUBAO_HTTP_BASE: HTTP base URL (default: “https://ark.cn-beijing.volces.com/api/v3”)

Features

• tts: Enable TTS (Text-to-Speech) API
• asr: Enable ASR (Automatic Speech Recognition) API
• chat: Enable Chat completion API
• embeddings: Enable Embeddings API
• images: Enable Image generation API
• tokenization: Enable Tokenization API
• full: Enable all features
• rustls: Use rustls for TLS (default)
• native-tls: Use native-tls for TLS

Modules

config

Configuration for Doubao API client.

error

Error types for Doubao API client.

spec

Type definitions for Doubao APIs.

Structs

Asrasr

ASR (Automatic Speech Recognition) API group.

Chatchat

Chat completions API.

Client

Doubao API client.

Embeddingsembeddings

Embeddings API.

Imagesimages

Images generation API.

Recognitionasr

File-based speech recognition API.

Speechtts

Speech synthesis API.

Streamingasr

Streaming speech recognition API.

StreamingSessionasr

A streaming recognition session.

Tokenizationtokenization

Tokenization API.

Ttstts

TTS (Text-to-Speech) API group.