Audoc
Hear the whole document — tables, charts and all.
Audoc turns PDFs, Word, PowerPoint, Excel, EPUB and more into structured listening sessions — narrating tables, chart trends and diagrams right where they appear, with on-device AI and offline neural voices. Live on Google Play; coming soon to iOS.
- Year
- 2026
- Platform
- Android
- Stack
- On-device Gemma 4 · Offline neural TTS · Document parsing · BYO-key AI
- Status
- Shipping
The vision
Read-aloud apps flatten a document into a stream of sentences and skip everything that isn't one — the table, the chart, the diagram that carried the point. Audoc is our bet that listening can explain more than plain text: keep the document's structure, describe each visual where it sits, and do it privately. Parsing, scripts, translation and voices run on the device by default; a cloud model is used only when you allow it, one document at a time. It is the same thesis as the rest of our work — capable AI, running on the device you already carry.
What's inside
Imports PDF, Word, PowerPoint, Excel, OpenDocument, EPUB, HTML, Markdown, CSV and images
Narrated and Plain Read modes that keep headings, tables, charts and diagrams in place
Tables read with headers and rows; charts described by trend, comparison, highs and lows
Local by default: optional on-device Gemma 4 for translation and visuals, offline neural voices
Automatic language detection, with matching voices for the original and the translation
Optional Gemini or OpenAI with your own key — cloud stays off per document until you allow it
Next project
K-MingleStart a project
Have a platform to plan or build?
Tell us the idea — a napkin sketch, a stalled roadmap, or a platform you'd rather hand to a senior team. We'll tell you honestly how we'd approach it.