Export formats

Plain text (TXT) — just the words, in readable paragraphs. Reach for it to read the transcript or paste it into a document. It carries no timing, so it's not for subtitles. Read more →

Markdown — the text with a metadata header and optional clickable timestamp headings. Best for note apps like Obsidian or Notion, where it drops in and stays linked to the video. The header only shows as fields in an app that reads Markdown frontmatter. Read more →

CSV — one row per segment, with start and end times and word counts. Made for spreadsheets and text analysis. It's data, not something you'd read top to bottom. Read more →

SRT — SubRip, the subtitle format almost every editor and player reads. Use it to add captions to a video. Its timestamps use a comma before the milliseconds — close to VTT but not interchangeable. Read more →

VTT — WebVTT, the subtitle format built for web video (the HTML5 <track> element). Use it for players on a web page. It looks like SRT but uses a dot before the milliseconds, so don't swap one file for the other. Read more →

JSON — the raw segments as structured data you can parse in code. Use it when you're building your own processing. This is the free, standard JSON — not the RAG one below. Read more →

RAG JSON — the transcript pre-split into overlapping chunks, each with a deep link (a link that jumps straight to that moment in the video), ready for a vector database (the store behind retrieval-augmented generation). Use it when you're loading transcripts into an AI search system. It's the one export that costs credits. Read more →