- Home
- AI Podcast Generators
- Audify
Audify
What it is
Audify is an open-source application that converts PDF, DOC, and DOCX files into editable two-speaker podcast scripts and downloadable MP3 episodes. It extracts document text and can use OCR when a scanned PDF contains little detected text. Users can review and change speaker turns, wording, and flow, choose host and guest voices, and select conversational, educational, or professional tones. Ollama is the primary script-generation path, with OpenAI as a fallback. OpenAI API access is required for text-to-speech, while Ollama is optional. The TTS service creates speech for each dialogue turn, combines segments into MP3, and provides playback and download. Audify uses FastAPI microservices with a React and Vite frontend and Docker Compose for local deployment. Its gateway API handles uploads, script generation, audio generation, job status, and MP3 downloads. Setup requires Docker Engine 24.x or later and Docker Compose Plugin v2 or later. Uploads are limited to 10 MB, and TTS accepts up to 100 dialogue turns. Jobs are stored in memory and reset when containers restart unless persistent storage is added.
For and against
For
- Imports PDF, DOC, and DOCX files
- Uses OCR for some scanned PDFs
- Allows dialogue editing before audio generation
- Exports downloadable MP3 episodes
- Available under the MIT license
Against
- Uploads are limited to 10 MB
- OpenAI API key is required for TTS
- Jobs reset after container restarts without persistence
- Maker says production-grade security controls are lacking
The verdict
Audify provides a clear document-to-podcast workflow with editable scripts and selectable voices. Its Docker-based deployment and in-memory jobs require technical setup and operational planning, especially for production use.
Who it suits
Audify suits users who want to turn office documents into editable audio conversations using a self-hosted setup. It is relevant to developers able to run Docker and configure Ollama or an OpenAI API key.
On the record
- Accuracy guidance
- The project advises users to review scripts, validate OCR output for scanned documents, and check generated audio and dialogue for factual accuracy and tone.github.com · read 4 Oct 2026
- API
- The gateway exposes endpoints for document upload, script and audio generation, job status, and MP3 download.github.com · read 8 Oct 2026
- Audio features
- The TTS service creates speech for each dialogue turn, mixes the segments into MP3, and provides voice options, status tracking, and downloadable audio.github.com · read 8 Oct 2026
- Audio generation
- The TTS service generates speech for dialogue turns, combines them into an MP3 podcast, and provides playback and download.github.com · read 4 Oct 2026
- Data handling
- The README says gateway and TTS job metadata use in-memory stores and jobs reset when containers restart unless persistent storage is added.github.com · read 8 Oct 2026
- Deployment
- The project is a FastAPI and React application orchestrated with Docker Compose, with setup instructions for running the stack locally.github.com · read 8 Oct 2026
- Document extraction
- It extracts text from PDF, DOC, and DOCX files and can use OCR for scanned PDFs when little text is detected.github.com · read 8 Oct 2026
- Document processing
- It extracts text from PDF, DOC, and DOCX files and uses OCR as a fallback for scanned PDFs with little detected text.github.com · read 4 Oct 2026
- Generation limit
- The TTS service's documented MAX_SCRIPT_LENGTH default is 100 dialogue turns.github.com · read 8 Oct 2026
- Integrations
- The documented external providers are Ollama for local model inference and OpenAI for TTS and optional LLM fallback.github.com · read 8 Oct 2026
- Intended use
- The project is described as an AI-assisted document-to-audio generator, and its disclaimer advises reviewing scripts and audio and checking OCR output.github.com · read 8 Oct 2026
- Job persistence
- Gateway and TTS job stores are in memory, so jobs reset when containers restart unless persistent storage is added.github.com · read 4 Oct 2026
- License
- The repository is licensed under the MIT License.github.com · read 8 Oct 2026
- Processing limit
- The TTS service configuration accepts up to 100 dialogue turns and allows up to five concurrent requests.github.com · read 4 Oct 2026
- Purpose
- Audify converts PDF, DOC, and DOCX documents into editable two-speaker podcast scripts and downloadable MP3 audio.github.com · read 8 Oct 2026
- Requirements
- The setup instructions require Docker Engine 24.x or later and Docker Compose Plugin v2 or later; an OpenAI API key is required for TTS, while Ollama is optional.github.com · read 4 Oct 2026
- Script editing
- Users can review the generated dialogue and edit speaker turns, wording, and flow before audio generation.github.com · read 4 Oct 2026
- Script generation
- Audify uses Ollama as the primary script-generation path and OpenAI as a fallback, with conversational, educational, and professional tone options.github.com · read 8 Oct 2026
