Podcast Chapter Markers: Formats and Transcript Workflow

A listener opens your episode in their podcast app and wants to find the section on pricing strategy at minute 34. Without chapters, they scrub blind. With chapters, they tap the title and land there in one step. That navigation difference is why chaptered episodes see stronger repeat-listen behavior than unchaptered ones on most shows that have tested the difference.

Three Chapter Formats in 2026

Chapters used to mean one thing: data embedded in the audio file. Now there are three distinct formats, each with different tradeoffs.

ID3 embedded chapters live inside the MP3 or AAC file itself as CHAP and CTOC frames. These are the traditional podcast chapter format, read natively by Apple Podcasts, Pocket Casts, Overcast, and Castro. The limitation: some podcast hosts strip ID3 tags during upload, and Spotify does not read them at all.

Description timestamps are plain-text chapters written directly into your episode notes. Format each line as a timestamp followed by a title: 00:00 Intro or 0:00 Intro. Apple Podcasts, Spotify, and YouTube all parse this format. It requires no audio editing, works with any podcast host, and is the easiest format to add retroactively after publishing. Spotify requires at least 3 chapters, each spaced at least 30 seconds apart, with titles under 40 characters.

Podcasting 2.0 JSON chapters are a separate JSON file hosted at a URL, linked from your RSS feed using the <podcast:chapters> tag. The file specifies chapter objects with startTime in seconds, an optional title, optional image URLs, and optional url links per chapter. The format (version 1.2) supports features that neither ID3 nor description timestamps can: per-chapter artwork and clickable links that appear during playback. Hosts like Buzzsprout, Transistor, and RSS.com support this format. Apps that read it include Apple Podcasts, Pocket Casts, Overcast, Fountain, and Podverse.

App Compatibility Table

The key correction from older guides: Spotify reads description timestamps and its own RSS-based chapter data, not ID3 CHAP frames. If Spotify matters to your audience, description timestamps are the only reliable path.

For YouTube, only description timestamps work. YouTube recognizes lines in the format 0:00 Chapter title or HH:MM:SS Chapter title, requires the first chapter to start at 0:00, and requires at least 3 chapters each at least 10 seconds long.

If you only implement one format, description timestamps cover the widest audience. If you use a hosting platform with native chapter support, doing both ID3 and JSON chapters is worth the small extra effort.

Apple's Automatic Chapter Generation

A significant change landed in November 2025 with iOS 26.2: Apple Podcasts now auto-generates chapters for English-language episodes that don't have any, labeling them "Automatically created" in the UI. Trailers and episodes under 10 minutes are excluded. Apple-supplied chapters are fine as a fallback, but auto-generated titles are generic. Shows that provide their own chapters retain full control over how topics are named.

You can disable Apple's automatic chapters in Apple Podcasts Connect. Any chapters you submit yourself override the auto-generated ones.

What Makes a Good Chapter Title

The point of a chapter is navigation, not decoration. Three rules that hold up in practice.

Keep it to 2-8 words. "Why revenue stalled in Q3" works. "The extended conversation we had about why our revenue growth was slower than expected in Q3" does not.

Specific beats generic. "Cold outreach teardown" is useful. "Sales tips" is not. Listeners scan chapter lists quickly; the specific title helps them find what they came for.

Match the actual content. If the chapter title says "fundraising advice" but the section is primarily about pitch deck design, listeners arrive at the wrong place and lose trust in the chapter labels.

How Many Chapters Per Episode

Chapter density that has worked well in practice:

• 15-25 minute episodes: 3-6 chapters
• 30-50 minute episodes: 6-10 chapters
• 60-90 minute episodes: 10-18 chapters
• 90+ minute episodes: 15-25 chapters

The target rhythm is a chapter break every 4-8 minutes. Too few chapters and listeners still scrub through long stretches. Too many and the chapter list becomes noise.

Generating Chapters From Your Transcript

Pre-2024, chapter generation was a manual job. You listened back, paused at each topic shift, and wrote a title. For a 60-minute episode, this took 30-60 minutes per episode.

Transcribing first and working from the transcript collapses that time to 5-10 minutes. The transcript makes topic boundaries visible on screen rather than audible on playback. You can scan a full hour of conversation in minutes, identify the 12-15 natural breaks, and write titles based on what's actually there.

The practical workflow:

1. Transcribe the episode. ConvertAudioToText handles audio and video files with speaker labels, which makes topic shift identification easier since you can see when the conversation changes direction.
2. Read through the transcript and mark the lines where topics clearly shift. These are your chapter boundary candidates.
3. Write a title for each section based on what's discussed, not what you intended to discuss.
4. Pull the timestamp for each marked line from the transcript.
5. Paste formatted chapters into your host's chapter editor or your episode description.

AI tools can suggest chapter boundaries directly from transcripts. The suggestions usually need editing: AI tends to over-segment (creating 25 chapters for a 40-minute episode) and sometimes under-names (using speaker quotes rather than topic titles). Treat AI output as a first draft, not a final product. Reviewing and editing AI chapter candidates still takes only 5-10 minutes per episode, versus building from scratch.

For the show notes that accompany chapters, the how to create podcast show notes automatically guide covers integrating transcript-based notes into the same workflow. The podcast show notes template has a structure that works well when chapters are listed as part of the episode description.

Tools That Embed ID3 Chapters

If you want ID3 chapters rather than just description timestamps, something has to write the CHAP frames into the audio file.

Podcast hosting platforms with built-in chapter editors include Buzzsprout (visual waveform editor plus plain-text Express Editor), Transistor, Captivate, Castos, and Podbean. You enter chapter titles and timestamps through their web UI, and the platform handles the embedding. For most independent podcasters, this is the lowest-friction option.

Buzzsprout's chapter editor is worth highlighting: the waveform editor lets you place chapters visually, while the Express Editor lets you paste chapters as plain text or JSON. Both produce ID3-embedded output. Buzzsprout also offers an optional AI chapter suggestion tool through its Cohost AI feature.

Digital audio workstations: Adobe Audition supports chapter markers on export. Audacity's support is limited and inconsistent across versions. If you're working in a DAW, check whether your host's web editor is simpler.

Standalone tools: Hindenburg Pro and a few other dedicated audio editors support chapter embedding. These add value if you're already using them for production, but are overkill for podcasters who only need the chapter step.

Common Mistakes That Reduce Chapter Usefulness

Vague titles. "Introduction", "Main Topic", "Outro" tell the listener nothing. A chapter list of generic labels is functionally useless. Every chapter should be specific enough that a listener who missed half the episode can guess what's in it.

Chapters that span 20+ minutes. Three chapters on a 90-minute episode still requires long scrubbing sessions inside each chapter. Aim for the density guidelines above.

Timestamps that drift after audio editing. If you add chapters before finishing edits, any cut or insert shifts every timestamp that follows. Always finalize chapters after the audio is locked. This is another argument for the transcript workflow: timestamps are generated from the final transcription of the finished file.

Forgetting Spotify's format. Adding ID3 chapters and skipping description timestamps means Spotify listeners get nothing. Since the description format is a paste job that takes 2 minutes, skipping it is a poor tradeoff.

Time Budget

For a weekly show, the transcript-based chapter workflow adds about 10-15 minutes per episode. Manual chapter creation (listening back and pausing) runs 30-60 minutes per episode.

Over a year of 52 episodes: roughly 9-13 hours with the transcript-based workflow versus 26-52 hours manually. The time saved is enough to record several additional episodes or meaningfully improve other parts of post-production.

For the broader post-production workflow that includes chapters, show notes, and transcript publishing, the transcription for podcasters complete guide covers the full sequence. If you're specifically looking to reduce the total time from recorded audio to published episode, the how to transcribe a 30-minute podcast quickly guide focuses on that time-compression problem directly.

If you just need a clean transcript to build chapters and show notes from, without a full editing suite, ConvertAudioToText uploads audio or video files and returns a timestamped transcript with speaker labels.

Common Questions

Which chapter format works across the most podcast apps?

Description timestamps (written into episode notes as plain text, one line per chapter) work on Apple Podcasts, Spotify, YouTube, and most other platforms without any file editing. If you only do one format, description timestamps are the path of least friction.

Does Spotify support ID3 chapter tags embedded in the audio file?

No. Spotify does not read ID3 CHAP frames. Spotify reads chapter timestamps from the episode description in Podlove format or its own RSS-based approach, and also auto-generates chapters from transcripts for English episodes. For Spotify, put your chapters in the episode description.

What is the Podcasting 2.0 JSON chapters format and which apps support it?

The Podcasting 2.0 JSON chapters format is a separate JSON file linked from your RSS feed via the podcast:chapters tag. Each chapter needs at minimum a startTime in seconds and optionally a title, image URL, and link. Hosting platforms including Buzzsprout, Transistor, and RSS.com support it. Apple Podcasts, Pocket Casts, Overcast, Fountain, and Podverse all read it. It is the only format that supports per-chapter images and links.

Does Apple Podcasts automatically generate chapters if I don't add them?

Yes, since iOS 26.2 (November 2025). For English-language episodes over 10 minutes, Apple Podcasts will generate chapters automatically when you haven't provided any. Auto-generated chapters are labeled "Automatically created" in the chapter list. You can disable this in Apple Podcasts Connect, and any chapters you supply yourself take priority.

Sources

• Podcasting 2.0 chapters specification
• Podcasting 2.0 JSON chapters format (jsonChapters.md)
• Apple Podcasts: automatic chapters FAQ (Podnews)
• Apple Podcasts for Creators: chapters and links
• Spotify: episode chapters
• ID3v2 Chapter Frame Addendum (id3.org)
• Pocket Casts: chapters support
• Buzzsprout: create chapter markers
• Buzzsprout: new chapter editor
• YouTube: video chapters
• Daring Fireball: Apple Podcasts AI-generated chapters