audia

Turn documents and ideas into audio, intelligently.

audia is an agentic Python package that converts PDFs — academic papers, reports, regulations — into podcast-style audio files. An LLM rewrites the content into natural spoken language (math in plain English, tables as sentences, citations removed) before passing it to a TTS engine.

Getting Started

• Installation
• Quick Start
• Configuration

Guides

• Pipeline
• CLI Reference
• Web UI
• TTS Backends
• LLM Backends
• Storage

API Reference

• audia — top-level package
• audia.config — Settings
• audia.agents — Pipeline agents
• audia.storage — Database models & access
• audia.cli — Command-line interface
• audia.ui — Web UI (FastAPI)

Project

• Changelog
• CHANGELOG

---

At a glance

Feature	Detail
LLM curation	Mandatory pass rewrites math, condenses tables, removes citations
TTS backends	edge-tts (free default), kokoro (local), OpenAI TTS
LLM backends	OpenAI, Anthropic
ArXiv research	Search → select → convert in one command
Voice input	Record a spoken query → STT → LLM distillation → ArXiv search
Web UI	FastAPI backend + React/Tailwind SPA
CLI	convert, research, listen, serve, info
Storage	SQLite via SQLAlchemy; papers + audio files
Python	3.10 – 3.13

Quick install

pip install audia