Skip to main content
If you have a transcript from a source CallVault doesn’t have a full connector for — or you just want to import one call without connecting an account — you can paste it directly. CallVault detects the format automatically and previews what it found before you save.

​
Importing a transcript

1

Open the paste import

Go to Import and choose Paste Transcript (or click Import Transcript from a workspace).
2

Paste a source link (optional)

If you have a share link, paste it in the Source link field. CallVault reads the link’s metadata — title, date, participants, thumbnail — and pulls in transcript text automatically for sources that support it.
3

Paste the transcript text, or choose a file

Paste the transcript into the text box, or click Choose transcript file to upload a .vtt, .srt, .txt, or .md file (up to 10MB).
4

Review the live preview

As you paste, CallVault parses the text with the same parser it uses server-side and shows a preview: detected title, date, attendees, speaker turns, and duration.
5

Adjust the details and save

Edit the title, date, attendees, or summary if needed, choose a destination workspace, and click Import Transcript.
Paste a link from any of these sources and CallVault automatically detects the format and, where possible, pulls metadata and transcript text:
Pasting a link isn’t the same as connecting an account. It’s a one-off import for a single call — for automatic ongoing sync, connect the source directly where CallVault supports it (see Fathom, Zoom, Fireflies, Read.ai, Grain, or Plaud).
If you paste a URL CallVault doesn’t recognize, it shows a warning and saves the link as metadata only — it won’t try to guess the format.

​
Supported transcript text formats

Paste the transcript text directly and CallVault auto-detects the format:
  • Fathom’s “Copy transcript” format — click Copy transcript in Fathom, then paste
  • Zoom VTT (WEBVTT) — paste raw VTT text or upload a .vtt file
  • SRT — subtitle-style timestamped transcripts
  • Otter.ai — Otter’s exported transcript text format
  • Loom / Grain timestamped transcripts
  • Fireflies / Read.ai / Calendly transcript text
  • Plain text — if the format isn’t recognized, CallVault saves the raw text and lets you fill in title, date, and attendees manually

​
What gets saved

Depending on what’s detected, the imported call includes:
  • Full transcript, parsed into speaker turns when the format supports it
  • Title, date, and attendees — pre-filled from the parsed content or link metadata, editable before saving
  • Summary — pulled from link metadata when available, editable
  • Duration — calculated from timestamps when present

​
Duplicate detection

If you paste a transcript that’s already been imported (matching source link or content), CallVault tells you it’s already in your vault instead of creating a duplicate, with a link to view the existing call.
Use Routing Rules with a “Source is” condition to automatically route pasted transcripts from a specific source into the right workspace.