Hey everyone,
I’ve been building a tool I wanted to share with the community.
Audio Brief transforms your Obsidian notes into full podcast-style audio conversations. Not text-to-speech — actual AI-generated dialogue between two hosts who analyze, debate, and riff on your content.
Demo: [embedded video or GIF]
Three modes available (+ one exclusive to the full system):
- The Boardroom — Succession-style strategic debate on your business docs (6-10 min)
- The Roast — Comedy podcast that tears apart your writing with love (4-10 min)
- The Briefing — Solo CIA-style intelligence brief, cold and precise (3-7 min)
Deep Trace — Intellectual detective duo dissecting your research (8-12 min) — available in the full Monolithos system
What I think makes it useful:
- Works directly with your vault files (select notes, hit TRANSMIT)
- Transcript auto-saves as .md back into your vault
- 10 languages with native-quality voices
- Audio player built into Obsidian — persistent playback across views
- You can type custom direction to guide the conversation before generating
Privacy note:
Only the content of your selected notes is sent for processing. Your vault structure, other files, and metadata are never accessed.
5 free generations (full Pro quality) to try it out. After that it’s $9.90/month.
Register for a free key at: Audio Brief — MONOLITHOS
Install via GitHub (Community Plugin review pending):
I built this as part of a larger project called Monolithos — a sovereign Life OS built on Obsidian. Audio Brief is the first module available as a standalone plugin.
Would love to hear your feedback. Especially interested in:
- Which mode you liked most
- What kind of notes worked best as source material
- Any bugs or rough edges
Thanks for being such a great community. Building on Obsidian has been a privilege.
— Silas