Bring your source
material.
Import EPUBs and TXT files or paste text straight in. Review extracted copy, metadata, and cover art before narration begins.
Star on GitHub
Turn books, articles, and anything in between into audiobooks that keep going—even when the work gets interrupted.
Awaaz breaks long text into durable synthesis jobs. It tracks every piece, stores audio checkpoints, and lets you return right where you left off.
See the flow ↘Import EPUBs and TXT files or paste text straight in. Review extracted copy, metadata, and cover art before narration begins.
Choose a backend, voice, pace, and chunking strategy. Preview the workload before you press generate.
Pause, retry, recover, and download when complete. Every chunk is visible, every job remains under your control.
Real books have chapters. Real systems have interruptions. Awaaz treats audio generation as a resumable process—not a fragile request you have to babysit.
The browser library holds your source files and active jobs in one place—while the API gives you the same controls in code.



Awaaz is a self-hosted, MIT-licensed project. Run it locally with Docker Compose, wire it into an existing pipeline, or point it at an OpenAI-compatible TTS backend.
$ docker compose up --build
Give your long-form content a durable voice—with Awaaz.