fireflies_get_transcript

Retrieve detailed transcripts from Fireflies with speaker analytics
READ

fireflies_get_transcript

Retrieves the complete transcript of a Fireflies meeting with speaker attribution and analytics. This is the primary tool for extracting meeting conversation data to create minds or analyze discussions.

Speaker Email Limitation

Important: Fireflies does not directly provide speaker email addresses. The MCP server infers emails by correlating speaker names with meeting attendee emails via name matching. For guaranteed accurate speaker emails, use Gong or Fathom.

Prerequisites

Fireflies Integration Required

This tool requires Fireflies integration to be configured. See the Fireflies Setup Guide for configuration instructions.

Parameters

transcriptId

Typestring
RequiredYes
DescriptionThe unique transcript identifier from fireflies_list_transcripts or fireflies_search_transcripts response
Exampletranscript_abc123

Request

1{
2 "transcriptId": "transcript_abc123"
3}

Response

1{
2 "transcriptId": "transcript_abc123",
3 "title": "Weekly Team Standup",
4 "date": "2024-11-15T10:00:00Z",
5 "duration": 1800,
6 "transcript": [
7 {
8 "speakerName": "Sarah Chen",
9 "speakerEmail": "sarah.chen@company.com",
10 "text": "Good morning team. Let's go around and share our updates for the week.",
11 "timestamp": "00:00:10",
12 "wordCount": 14,
13 "speakingTime": 5
14 },
15 {
16 "speakerName": "Mike Johnson",
17 "speakerEmail": "mike.johnson@company.com",
18 "text": "I finished the API integration we discussed last week. It's ready for QA testing now.",
19 "timestamp": "00:00:18",
20 "wordCount": 16,
21 "speakingTime": 6
22 },
23 {
24 "speakerName": "Alex Rivera",
25 "speakerEmail": "alex.rivera@company.com",
26 "text": "Great work Mike. I'll start testing that this afternoon. My update is I'm working on the database migration.",
27 "timestamp": "00:00:28",
28 "wordCount": 20,
29 "speakingTime": 8
30 }
31 ],
32 "speakerAnalytics": [
33 {
34 "name": "Sarah Chen",
35 "email": "sarah.chen@company.com",
36 "totalSpeakingTime": 360,
37 "wordCount": 720,
38 "talkRatio": 0.20
39 },
40 {
41 "name": "Mike Johnson",
42 "email": "mike.johnson@company.com",
43 "totalSpeakingTime": 540,
44 "wordCount": 1080,
45 "talkRatio": 0.30
46 },
47 {
48 "name": "Alex Rivera",
49 "email": "alex.rivera@company.com",
50 "totalSpeakingTime": 900,
51 "wordCount": 1800,
52 "talkRatio": 0.50
53 }
54 ]
55}

Response Properties

PropertyTypeDescription
transcriptIdstringThe transcript identifier
titlestringMeeting title from calendar
datestringISO 8601 timestamp of the meeting
durationnumberMeeting duration in seconds
transcriptarrayArray of transcript segments in chronological order
transcript[].speakerNamestringSpeaker’s name
transcript[].speakerEmailstringInferred email (via name matching with attendees)
transcript[].textstringThe spoken text for this segment
transcript[].timestampstringTimestamp within the meeting (HH:MM:SS format)
transcript[].wordCountnumberNumber of words in this segment
transcript[].speakingTimenumberDuration of this segment in seconds
speakerAnalyticsarrayAggregate analytics per speaker
speakerAnalytics[].namestringSpeaker’s name
speakerAnalytics[].emailstringInferred email address
speakerAnalytics[].totalSpeakingTimenumberTotal speaking time in seconds
speakerAnalytics[].wordCountnumberTotal words spoken
speakerAnalytics[].talkRationumberPercentage of total speaking time (0-1)

Common Use Cases

Extract conversation data to train digital minds on team member communication styles.

Example:

"Get the transcript from my standup meeting and create
minds for each team member"

Natural language workflow:

  1. Uses fireflies_list_transcripts or fireflies_search_transcripts to find standup
  2. Uses fireflies_get_transcript to retrieve full conversation
  3. Identifies unique speakers from analytics
  4. Creates mind for each speaker
  5. Uploads their transcript segments as training data

What you get:

  • Minds trained on actual communication patterns
  • Speaker analytics for each participant
  • Ability to simulate future discussions
  • Note: Verify speaker emails due to name-matching inference

Use speaker analytics to understand meeting dynamics.

Example:

"Get speaker analytics from my leadership meetings
this month and show me participation patterns"

Natural language workflow:

  • Lists leadership meetings
  • Retrieves transcripts with analytics
  • Analyzes talk ratios and speaking time
  • Identifies participation patterns

What you get:

  • Detailed speaker metrics
  • Talk time distribution
  • Word count analysis
  • Participation insights

Extract high-quality conversation segments for team training.

Example:

"Get transcripts from my successful client calls and
extract the discovery question patterns"

Natural language workflow:

  • Identifies successful client calls
  • Retrieves transcripts
  • Analyzes question patterns
  • Extracts effective techniques

What you get:

  • Real-world conversation examples
  • Proven communication techniques
  • Training materials from actual meetings
  • Best practice documentation

Error Responses

1{
2 "error": "Not Found",
3 "message": "Transcript with ID 'transcript_abc123' does not exist"
4}

Solution: Verify the transcript ID is correct. Use fireflies_list_transcripts or fireflies_search_transcripts to get valid IDs.

1{
2 "error": "Unauthorized",
3 "message": "Fireflies authentication failed"
4}

Solution: Verify your Fireflies API key is configured correctly.

1{
2 "error": "Forbidden",
3 "message": "You do not have permission to access this transcript"
4}

Solution: Check your Fireflies permissions for this meeting.

1{
2 "error": "Bad Request",
3 "message": "Transcript not yet available"
4}

Solution: Meeting may still be processing. Wait and retry. Check status in Fireflies.

Best Practices

Verify Speaker Emails
  • Critical: Speaker emails are inferred, not guaranteed
  • Review speaker attribution before creating minds
  • Manually verify emails for important use cases
  • Consider using Gong or Fathom if accurate emails are required
  • Check that speaker names match attendee names reasonably well
  • Use talk ratios to find balanced conversations
  • Identify dominant vs. quiet participants
  • Select high-engagement meetings for training
  • Track participation trends over time
  • Filter to specific speakers when creating individual minds
  • Use word count to identify substantial segments
  • Combine transcripts from multiple meetings for richer training
  • Preserve timestamps for temporal analysis

Integration with Mind Reasoner

Creating Minds from Fireflies Transcripts

1

Find Meetings

Use fireflies_list_transcripts or fireflies_search_transcripts to find relevant meetings

2

Get Transcript

Use fireflies_get_transcript to retrieve the full conversation with analytics

3

Verify Speaker Emails

Review inferred speaker emails for accuracy (critical step!)

4

Extract Speaker Segments

Filter transcript to segments from the target speaker

5

Create Mind

Use Mind Reasoner’s create_mind with speaker details

6

Upload & Train

Upload filtered transcript and create snapshot

Natural language example:

"Find my team meetings from last quarter and create minds
for each team member based on their participation"

Important: The AI will attempt to infer speaker emails, but you should verify the attribution is correct before using minds in production.

See Also