Audio Chapters: Turn Long Recordings Into Navigable Sections
Some recordings are long because they're important. Most are long because humans love to talk.
Either way, a 90-minute file shouldn't feel like a 90-minute task.
Transcribe.so automatically generates chapters and sections so you can navigate long audio like a document.
How Chapter Generation Works
- Upload your audio — Interviews, podcasts, meetings, lectures
- AI analyzes the content — Identifies section shifts and key moments
- Get a chapter outline — Jump to any section instantly
No manual timestamps. No scrubbing through the timeline.
What Chapters Give You
- Instant navigation — Click a chapter, jump to the moment
- Skimmable outline — Know what's in a recording before you commit to listening
- Shareable sections — Send "the part that matters" without a 90-minute download
- Section grouping — Related ideas cluster together
When Chapter Generation Is a Superpower
| Use Case | Without Chapters | With Chapters |
|---|---|---|
| Interviews | Scrub to find the answer | Click "Q3: Pricing discussion" |
| Lectures | Re-watch the whole thing | Jump to "Section: Neural Networks" |
| Podcasts | Guess where the good part is | See the outline, pick your segment |
| Training | Skip around randomly | Navigate by module |
Pair Chapters With Other Features
- Semantic search — Find ideas even if you don't remember the exact words. Chapters pair well with subtitle export — jump to a chapter and export subtitle cues for that section.
- AI Q&A — Ask "what was the main point?" and get a summary with timestamps
- Speaker ID — Know who's talking in each chapter
Frequently Asked Questions
How many chapters are generated?
It depends on the content. A 1-hour interview might get 8-15 chapters. A monologue with few section changes might get fewer.
Can I edit or rename chapters?
Currently, chapters are AI-generated. Manual editing is not available yet.
Does this work for any audio?
Yes. Interviews, podcasts, meetings, lectures, webinars, and more. You can also request chapter generation via the ChatGPT Custom GPT or the Claude Custom Connector without opening the dashboard.
What's the maximum length?
Uploads support up to ~8 hours of audio (configurable) with a 500MB per-file limit. For 8+ hour files, Qwen3-ASR-Flash supports up to 12 hours natively.
Stop drowning in long audio. Generate chapters automatically →