Transcript export2 min read
Export a NotebookLM transcript as SRT, VTT, Markdown, or CSV
NotebookLM gives you a video and a chat panel. It does not give you a transcript file. Once captions exist, choosing the right export format turns a ten minute overview into study material you can search, edit, and reuse.
Pick the format by what you will do next
| If you want to | Export | Why |
|---|---|---|
| Edit the video in Premiere, DaVinci, or CapCut | .SRT | The universal subtitle interchange format with numeric cue indices. |
| Publish captions on the web | .VTT | WebVTT supports cue positioning and styling in HTML5 players. |
| Read or print the transcript | .TXT | Plain paragraphs with optional timestamps, readable anywhere. |
| Build study notes or an Obsidian vault | .MD | Markdown headings, timestamps, and bookmarked moments preserved. |
| Build an Anki deck or spreadsheet | .CSV | One row per cue with start, end, speaker, and text columns. |
| Feed a script or downstream tool | .JSON | Structured cues with provider, model, and language metadata. |
How timing offsets are applied
If captions feel consistently early or late, shift them with the bracket keys before exporting. The offset you set is baked into every cue in the exported file, so the .SRT you hand to an editor matches what you saw on screen. Reset with the backslash key.
Speaker labels and bookmarks
Audio Overview podcasts have two hosts, and LM Subs keeps those labels when the provider reports them. Bookmark any line while watching and the Markdown export collects your bookmarked moments into a review section at the bottom of the file.
Copy a single spoken line
Clicking the caption copies the active line with its timestamp, which is the fastest way to quote a specific moment. From the transcript drawer you can also insert a formatted quote straight into the NotebookLM chat panel to ask a follow-up question about it.
Frequently asked questions
Can I export captions without installing anything?
No. NotebookLM does not expose a transcript download for Video Overviews, so captions have to be generated from the audio track first. LM Subs generates them in the browser and then exports in six formats.
Do exported files include the speaker names?
Yes, when the speech-to-text provider returns speaker or segment labels and speaker labels are enabled in settings. You can toggle them off for a clean caption track.
Can I re-import an edited subtitle file?
Yes. Use the Import control in the caption options to load an edited .SRT, .VTT, or .JSON file back into the player, and it is saved to your transcript library.