Audio

Overview

NovaAI Audio combines Speech-to-Text (STT) and Text-to-Speech (TTS) capabilities into a single API. Whether you're building voice assistants, meeting transcription tools, accessibility features, podcasts, or customer support systems, the Audio API provides fast, accurate, and scalable speech processing.

The same authentication, request structure, and SDKs used throughout NovaAI also apply to audio workloads.

Audio Capabilities

Audio Processing Workflow

Speech-to-Text

Convert recorded speech into structured text.

1import { NovaAI } from "@novaai/sdk";
2 
3const client = new NovaAI({
4  apiKey: process.env.NOVA_API_KEY,
5});
6 
7const transcript = await client.audio.transcribe({
8  model: "nova-audio-1",
9  file: "./meeting.mp3"
10});
11 
12console.log(transcript.text);

Text-to-Speech

Generate natural voice output from text.

1const speech = await client.audio.speech({
2  model: "nova-audio-1",
3  voice: "nova",
4  input: "Welcome to NovaAI documentation."
5});
6 
7console.log(speech.url);

Request Properties

Audio model identifier.

Audio file to transcribe.

Spoken language.

Include timestamps in the response.Speech-to-Text

Text-to-Speech

Audio generation model.

Voice preset used for synthesis.

Text to convert into speech.

Output audio format.

Supported Audio Formats

Response Example

Speech-to-Text

{
  "id": "audio_91ab72",
  "text": "Welcome to the NovaAI documentation.",
  "language": "en",
  "duration": 4.2
}

Voice Preview

Available Voices

Best Practices

• Upload high-quality recordings for better transcription accuracy.

• Minimize background noise whenever possible.

• Choose a voice that matches your application's personality.

• Use compressed formats like MP3 for faster uploads.

• Split long recordings into smaller segments for improved processing.

Common Errors

Unsupported Audio Format

The uploaded file format is not supported.

Solution

Convert the file to MP3, WAV, M4A, FLAC, or OGG before uploading.

Corrupted Audio File

The uploaded file cannot be decoded.

Solution

Verify the audio file and upload a valid recording.

Audio Duration Exceeded

The uploaded recording exceeds the maximum supported duration.

Solution

Split the recording into smaller segments and process them individually.

Frequently Asked Questions

Next Step

Now that you've learned how NovaAI processes speech, continue to Vision to explore image understanding, visual reasoning, and multimodal AI capabilities.