Plain text (TXT)
The TXT export gives you the transcript as plain text in two variants: continuous paragraphs, or one line per segment — each short, timed chunk the transcript is split into — prefixed with an HH:MM:SS (hours:minutes:seconds) timestamp. It is the only format available without an account.
Without timestamps (paragraphs)
Segments are merged into paragraphs. A new paragraph starts when there is a gap of more than 2 seconds between segments, when a paragraph passes about 90 seconds, or when the previous segment ends on ., ! or ?.
Welcome to this short introduction to vector databases.
A vector database stores embeddings and retrieves them by similarity.
That makes it the backbone of most retrieval augmented generation systems.With timestamps
Each segment is one line: an HH:MM:SS timestamp of the segment's start, then two spaces, then the text.
00:00:00 Welcome to this short introduction to vector databases.
00:00:03 A vector database stores embeddings and retrieves them by similarity.
00:00:07 That makes it the backbone of most retrieval augmented generation systems.Availability
TXT (both variants) is the only export you can download — or copy — without an account. The other six formats need a free account.
Sources
- INDXR (own code) — TXT paragraph rules, timestamp formatVerified against packages/shared/src/utils/formatTranscript.ts (generateTxt, createParagraphMode)



