Crate vtt_rs 

Voice-to-Text Streaming Library

This library provides real-time audio transcription capabilities for AI agents and other applications that need situational awareness through speech recognition.

Architecture

The library is organized into several key components:

• TranscriptionService: Main service that orchestrates audio capture and transcription
• Config: Configuration for chunk duration, API endpoint, and model selection
• TranscriptionEvent: Events emitted for successful transcriptions and errors

Audio is captured from the default input device, chunked into configurable segments, and sent to OpenAI-compatible transcription APIs. Results are delivered via an event-driven async channel.

Examples

Basic Usage

use vtt_rs::{TranscriptionService, Config, TranscriptionEvent};

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    let config = Config::default();
    let api_key = std::env::var("OPENAI_API_KEY")?;

    let mut service = TranscriptionService::new(config, api_key)?;

    // Start listening and transcribing
    let (mut receiver, _stream) = service.start().await?;

    // Process transcription events
    while let Some(event) = receiver.recv().await {
        match event {
            TranscriptionEvent::Transcription { chunk_id, text } => {
                println!("Heard: {}", text);
            }
            TranscriptionEvent::Error { chunk_id, error } => {
                eprintln!("Error: {}", error);
            }
        }
    }

    Ok(())
}

Custom Configuration

use vtt_rs::{Config, TranscriptionService};
use std::path::PathBuf;

let config = Config {
    chunk_duration_secs: 3,
    model: "whisper-1".to_string(),
    endpoint: "https://api.openai.com/v1/audio/transcriptions".to_string(),
    out_file: Some(PathBuf::from("transcripts.log")),
};

let api_key = std::env::var("OPENAI_API_KEY")?;
let mut service = TranscriptionService::new(config, api_key)?;

Structs

Config

Configuration for the transcription service.

OnDeviceConfig

Configuration for the on-device Whisper backend.

TranscriptionService

The main transcription service.

Enums

TranscriptionEvent

Events emitted by the transcription service.

Traits

Context

Provides the context method for Result.

Type Aliases

Result

Result<T, Error>