Chapters

The Chapters node lists the chapters of a podcast or interview recording in one step. It transcribes the recording and finds its chapters, each with a title and a start time, ready to paste into show notes or a video description. Nothing is rendered: the result is text.

Inside, the node runs the same steps that you can wire by hand: Transcribe with speaker labels and word timings, then Edit Plan in Chapters mode. Each step runs as its own job, so a long recording, a cancel or a failure behaves as it does in a hand-built workflow.

Chapters runs its planning step on Nodaro Cloud. On a self-hosted install, connect the install to Nodaro Cloud, under Integrations › nodaro.ai or with an API key. Without a connection, the node is not offered and a run is refused before anything starts.

When to use it

• Add chapter markers to the show notes of a podcast episode.
• Add timestamps to the description of a video.
• Make a table of contents for a long interview or lecture.
• Pass the chapter list as JSON to a player, a website or a later step of the workflow.

Use Edit Plan in Chapters mode yourself when you already have a transcript. To tighten the episode as well, use Tighten Episode.

Quick start

Inputs

A recording that is longer than 3 hours, or a recording with no sound, is refused before anything is charged. With more than one recording wired in, a run reads the first one.

Outputs

How the show notes are written:

• The first chapter starts at 0:00, where a listener's player starts.
• Hours appear only when they are needed. Once a chapter starts at 1:00:00 or later, every line uses hours: 0:00:00 Intro, 1:04:10 Questions. A 70-minute episode whose last chapter starts at 58:00 keeps the short form.
• The times only go up. When two chapters start in the same second, the first one is kept and the others are left out. A chapter that starts within the first second is left out too.
• A title is one line. Line breaks and runs of spaces inside a title become one space.

In a published app, the show notes are shown as text.

Settings

If a run fails

The steps that finished stay finished. The transcript is kept, and it is not charged again. Press Continue on the node: only the step that failed and the steps after it run again.

A node fed by a List runs once per item. It has no Continue button, so run it again instead.

When the plan finds no chapters, the run fails with the message "The plan found no chapters in this recording." The per-run fee is refunded. The transcript and the plan ran, so they are charged at their own rates. Every step finished, so the run cannot be continued and the node offers no Continue. Run the node again, with other instructions or another recording.

Credits

Chapters charges each step at its own rate, the same as the node it stands for, plus a small per-run fee. The node shows the full estimate, fee included, before you run.

• The per-run fee is charged once per run. A failed run's fee is refunded, and Continue holds it again.
• A List runs each item as its own run, with its own fee.
• The estimate counts the recording's minutes the way Edit Plan does. A recording with no known length is estimated at the 180-minute ceiling. How Edit Plan counts the minutes is explained on Edit Plan.

Tips

• Say how many chapters you want. Use Instructions for the number of chapters and the style of the titles, for example titles under six words.
• Use the right output. Wire Show notes into a text field or a publishing step. Wire Chapters (JSON) where you need the start times as numbers.
• Draft on Economy. Economy is the default. Try a higher Planning quality when the chapters need to be more precise.

Troubleshooting

From the API

POST /v1/episode-chapters starts a run and answers 202 with jobId and executionId. The body takes mediaUrls (one recording URL), and optionally planTier, instructions and styleGuide. Poll the job with GET /v1/jobs/:id/status.

• The job's output_data carries the show notes (text), the chapter list (json), the plan as Edit Plan made it (plan) and the transcript (transcript). A failed run keeps what completed and names the failed step in failedStep.
• An Idempotency-Key header makes a retry of the same request answer with the same jobId and "deduped": true, and start nothing.
• POST /v1/episode-chapters/estimate takes the same body and answers { data: { totalCredits, stepsCredits, feeCredits, durationSec } }, without starting anything.
• POST /v1/episode-chapters/continue with { jobId, "mode": "from-failed" } runs what a failed run did not complete. A run whose every step finished answers 409 continuation_not_failed. Chapters renders nothing, so "mode": "render-final" answers render_final_not_preview.

The SDK (client.edit.episodeChapters, client.edit.estimateEpisodeChapters, client.edit.continueUseCase), the CLI (nodaro edit episode-chapters, with --estimate for a quote, and nodaro edit continue-use-case) and the MCP tools make_chapters, estimate_use_case and continue_use_case call the same routes. See Run a single node, Editing in the SDK and the MCP analysis and EDL tools.