HomeFeaturesAdd Chapter Markers
Add Chapter Markers

Chapters Written by
the Content, Not by You.

The transcript is analysed for topic shifts and long pauses. Chapter markers with descriptive titles are placed at every natural boundary, exportable to YouTube chapters, Spotify chapters and podcast show notes.

Premiere Pro Final Cut Pro DaVinci Resolve YouTube · Spotify · Podcasts
How Topic Detection Works

Chapters where the content actually changes.

The chapter detection algorithm analyses the transcript using a sliding window approach. It measures the semantic similarity between adjacent transcript segments using keyword overlap and topic vocabulary. When the similarity drops below a threshold — indicating a genuine topic change — a chapter boundary is placed at that point.

Chapter boundaries are also confirmed by long pauses in the audio, since speakers naturally pause when transitioning between major topics. The two signals — semantic shift and audio pause — are combined to produce boundary candidates with a confidence score.

  • Semantic similarity scoring between adjacent segments
  • Audio pause confirmation at detected boundaries
  • Topic keyword extraction for chapter title generation
  • Minimum chapter length prevents too-short chapters
  • You edit chapter titles in the panel before placing markers
Generated Chapters
0:00 Introduction 97%
4:22 The Problem With Manual Editing 91%
12:47 How AI Transcription Works 88%
21:05 Quality Control in Post 93%
38:14 Pricing and Next Steps 85%
Minimum Spacing Control

Set chapter density for your format.

The minimum chapter duration setting prevents chapters from being too short. YouTube recommends chapters be at least 10 seconds long to appear in the chapter list. Podcast platforms typically work better with chapters spaced 3–10 minutes apart. EditLint AI respects your minimum spacing setting when selecting which boundaries to keep.

  • YouTube short-form — minimum 10 seconds, 4–8 chapters typical
  • YouTube long-form — minimum 2 minutes, 6–12 chapters typical
  • Podcast — minimum 3 minutes, chapters every major topic shift
  • Custom — set any minimum from 10 seconds to 60 minutes
Export Formats

One click to YouTube, Spotify and show notes.

After placing chapter markers on your timeline, export the chapter list in any of the supported formats from the EditLint panel.

  • YouTube chapters — timecoded text list in the format YouTube recognises for the video description (e.g. 0:00 Introduction)
  • Spotify chapters — JSON format for Spotify Podcaster chapter import
  • Podcast show notes — Markdown formatted chapter list for Substack, Ghost and most podcast CMS platforms
  • CSV — plain timecode and title table for any other workflow
How It Works

Transcribe. Detect. Title. Export.

1

Transcribe and Segment

Local Whisper transcription produces the full transcript. The transcript is split into segments based on natural sentence boundaries. Each segment is represented as a bag of keywords for semantic similarity comparison.

2

Detect Boundaries and Generate Titles

Adjacent segments are compared for semantic similarity. Boundaries where similarity drops significantly, confirmed by a long pause in the audio, are selected as chapter start points. For each chapter, the most prominent keyword phrases from the chapter content are used to auto-generate a title. You can edit any title before placing markers.

3

Place Markers and Export

Click Place Markers to write chapter markers to your NLE timeline. Chapter marker names include the generated title so they're immediately readable in your NLE's marker panel. Click Export to generate the chapter list in your chosen format for copying into your video description, podcast CMS or Spotify.

Chapters done before you finish the edit.

AI-detected topic breaks with generated titles. Exportable to YouTube, Spotify and show notes. Zero manual timestamp copying.

No credit card required · Works inside your NLE · Cancel any time