Audify Documentation

Convert ebooks and PDFs to audiobooks using AI text-to-speech.

Getting Started

• Installation
  • Requirements
  • Install from PyPI
  • Install from Source
  • Verify Installation
  • Docker Setup (Optional)
• Quick Start
  • Option 1: Cloud TTS (Easiest)
  • Option 2: Kokoro TTS (Local, Free)
  • Option 3: Qwen-TTS (Local, Free)
  • Next Steps

User Guide

• Usage
  • Basic TTS Conversion
  • LLM-Powered Audiobook Generation
  • Task System
  • Two-Stage Workflow (Process / Synthesize)
  • Directory Processing
  • Listing Available Options
  • Output Structure
• Task System
  • Built-in Tasks
  • Custom Prompt Files
  • Writing Effective Prompts
  • Programmatic Task Registration
• Configuration
  • The .keys File
  • TTS Provider Comparison
  • Google Cloud TTS Configuration
  • Docker Services
  • Supported Languages
  • Pipeline Environment Variables
• REST API
  • Starting the API
  • Endpoints
  • Examples

Reference

• CLI Reference
  • Synopsis
  • audify [OPTIONS] [PATH]
  • audify list-tasks
  • audify validate-prompt
• API Reference
  • CLI entry point (audify.cli)
  • Conversion factory (audify.convert)
  • Audiobook creator (audify.audiobook_creator)
  • Text-to-speech synthesizers (audify.text_to_speech)
  • REST API (audify.api.app)
  • Readers
  • Prompt system
  • Utilities
• Architecture
  • Overview
  • Pipeline
  • External Services
• Using Commercial APIs with Audify
  • Supported APIs
  • Setup
  • Usage
  • Complete Example
  • Comparing Ollama vs Commercial APIs
  • Model Recommendations
  • Troubleshooting
  • Environment Variables (Alternative to .keys file)
  • Cost Considerations
  • Technical Details

Design & Architecture Decisions

• QA Pipeline Graph
  • Cycle 1 — text-quality escalation edge
  • Cycle 2 — script-validity reroute edge
  • Cycle 3 — fidelity retry edge
  • Sub-graphs
• Quality assurance is a cyclic graph, not a central LLM agent
  • Status
  • Considered options
  • Consequences
• LangGraph as the cyclic-graph substrate
  • Status
  • Considered options
  • Consequences
  • Skeleton scope (deferred behaviours)

Development

• Contributing
  • Development Setup
  • Development Workflow
  • Task Commands
  • Code Quality
  • Project Structure
• Changelog
  • [Unreleased]
  • [0.1.0] - 2024-01-01

Troubleshooting

• Troubleshooting: No Audio Files Created
  • Common Symptom
  • Quick Diagnosis
  • Root Causes & Solutions
  • Detailed Debugging
  • Log Files
  • Testing TTS Synthesis Directly
  • Environment Variables
  • Getting Help
  • Quick Reference: Working Setup
  • Still Having Issues?

What is Audify?

Audify is a pipeline and CLI tool that transforms written content (EPUB, PDF, TXT, MD) into high-quality audiobooks using:

• Multiple TTS Providers – Kokoro (local), Qwen-TTS (local), OpenAI, AWS Polly, Google Cloud TTS

• LLM-powered audio generation – Transform text into audiobook scripts, podcasts, lectures, summaries, and more

• Flexible task system – Use built-in prompts or provide your own custom prompts

Quick Example

# Install
pip install audify-cli

# Convert an EPUB to audiobook (direct TTS)
audify book.epub --task direct

# Generate an LLM-enhanced audiobook
audify book.epub --task audiobook

# Use a different task (podcast-style)
audify book.epub --task podcast

# Use a custom prompt
audify book.epub --prompt-file my-prompt.txt

Indices and tables

• Index

• Module Index

• Search Page