FAQ
Common questions about supported formats, privacy, and output.
Frequently asked questions (7)
Is my transcript uploaded or stored on a server?
No. Timestamp parsing and every output conversion run entirely in your browser with JavaScript. Your pasted transcript is never uploaded or saved to any server, and the tool makes no outbound network requests with it. For convenience, only your latest session is kept in this browser’s localStorage so you can continue after a refresh.
Which timestamp formats are recognized?
Bracketed [00:01:23], short MM:SS like 00:01, H:MM:SS like 1:23:45, and parenthesised (00:01:23) are all supported. SRT/VTT millisecond suffixes (.500 or ,500) and cue ranges such as 00:00:05,000 --> 00:00:08,000 are handled too, using the start time. Mixed formats in one document are parsed line by line automatically.
Does it work with CLOVA Note exports?
Yes. CLOVA Note copies text as name-then-timestamp, like "Speaker 1 00:01", and the formatter splits that into speaker and time automatically. Select all in the note view, copy, paste it in, and the lines are grouped by speaker and converted into whichever output tab you pick: Markdown, Notion, plain text, or YouTube chapters.
Which tab should I use for Notion vs Word?
Use the Notion tab for Notion: it omits Markdown emphasis and separates speaker/timecode headers from body lines, so Notion auto-creates blocks on paste. Use the Plain (Word) tab for Word, Google Docs, or email — it produces clean "[00:00] Speaker: text" lines that paste cleanly anywhere without stray symbols.
How are YouTube chapters generated?
The YouTube chapters tab treats each speaker change as a chapter start and builds a "00:00 Title" list. Because YouTube requires the first chapter to be 00:00, the first line is normalized automatically. Copy the list into your video description to apply chapters. If there are no speaker labels, the start of each line is used as the title.
Does filler removal damage the actual speech?
It only strips hesitation words at the very start of a line and preserves the real content. For example, "um so let us begin" becomes "so let us begin". Removal is limited to short fillers, and letters inside words are never touched. Leave the option off to keep the source text exactly as pasted.
How do time-based sections work?
Pick 5, 10, or 15 minutes and the note is split into sections with time headers. Sections are computed first and speakers are grouped within each one, so even a single continuous speaker is split at a section boundary. For short meetings, "No split" usually reads best.