Evidence · first-party tested/Best AI Meeting Notetakers for Accurate Transcripts, Summaries, and Action Items
Notta’s transcript capture was accurate on the evaluated standup: the report says it correctly captured names, tool names, numbers, and engineering jargon with no significant word-level errors, silent hallucinations, or misheard terms.
What was measured
Transcription Accuracy
Word accuracy on the shared call, especially names, tools, numbers, and jargon.
decisive for this rankingtransformation
If the transcript gets names, numbers, and jargon wrong, the note-taker has failed at the core job of capturing the call accurately. (3 of 3 judges)
What was given, what came back
Test input: AI Demos Daily Standup — 31 July 2026 · image · group: ai-meeting-notetaker
Input — what we sent
AI Demos Daily Standup — 31 July 2026
A real 25-minute technical engineering daily standup with 14 attendees and about 10 active speakers, used as the single parallel-capture meeting for evaluating AI meeting notetakers on transcription, diarization, summaries, action items, search/chat, and collaboration features.
Why this input is hard
- · Transcription accuracy for real names, tool names, numbers, and technical jargon
- · Speaker diarization across multiple active speakers
- · Robustness to overlapping speech, crosstalk, and rapid turn-taking
- · Join reliability for bot-based and botless capture
- · Summary quality on identical source material
- · Action-item extraction with correct owners and commitments
- · Topic segmentation of standup updates
- · Search and chat grounded in the meeting content
- · Sharing, API, MCP, integrations, plan limits, languages, and privacy feature coverage
Output — unretouched
Also checked on this input — same tool, 8 other criteria
Action-Item Extraction✓ WorkedThe action-item list extracted the real commitments from the call, formatted them as checkbox items with @mentions, and the report says owner assignment was correct with no false positives.Chat with Notes / Ask Questions✓ WorkedThe Q&A interface answered a natural-language question with a grounded response from the meeting record, including the specific date "6th August," and the report observed no hallucinations.Join Method & Reliability✓ WorkedThe bot-based Google Meet join was reliable in the tested call: Notta Bot appeared in the meeting list, admitted/managed normally, and the capture ran through the end of the session without disconnects or plan-limit cutoffs.Search Across Notes◐ MixedSearch works inside a meeting transcript through AI Chat and returns exact timestamps in plain text, but the timestamps are not clickable, and the report says this was not tested across meetings.Speaker Diarization⚠ StruggledThe transcript contained a line labeled with another notetaker’s name (HappyScribe), which indicates cross-tool contamination or labeling error and breaks speaker attribution for that segment.Speaker Diarization◐ MixedSpeaker attribution was mostly correct, with nearly all speakers identified by name, but the transcript still showed some misattributed lines, so diarization was not fully reliable for every turn.Summary Quality✓ WorkedThe generated meeting summary was comprehensive and skimmable, with structured sections such as Task & Issue Management and a mindmap-style organization that reflected the meeting flow.Topic Segmentation✓ WorkedThe meeting was segmented into useful topic blocks rather than one blob; the report names three sections, including Task & Issue Management, Individual Progress Updates, and API Benchmarking Task.
Provenance
- Observation
- 893f5543-bd03-4e2a-ba30-f6426940628b
- Evidence run
- ace58582-3d1e-48ee-996c-9b3cd03f27a2
- Study
- AI Meeting Notetakers — Capture Accurate Transcripts, Summaries & Action Items From Live Calls
- Research task
- 86baxegnv
- Tested at
- not recorded
- Source
- first-party
- Evidence state
- verified
- Proof shown
- input + output shown
- Cost / latency
- not captured
- Repeat run
- not captured
- Tester
- not captured
The last three rows are honest blanks, not placeholders — our capture has no field for them yet.
Query this
get_evidence({
tool: "notta",
scenario: "ai-meeting-notetaker"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, same check — 7 other tools
measured on Transcription Accuracy
Fathom◐ MixedFathom's transcript mostly preserves the meeting's names and technical content, but the report records one confirmed name-level error: "Mahreen" was rendered as "Meryl."Fellow✓ WorkedThe transcript was near-clean: the tool captured nearly all names, technical jargon, and numbers correctly, with no significant misheard terms or hallucinations observed in the tested meeting.Fireflies.ai✓ WorkedGenerated a timestamped transcript view, and the report says the full transcript was very accurate: nearly all names, tools, jargon, and numbers were captured correctly with no significant misheard terms or hallucinations.Granola✗ FailedOn this 25-minute, multi-speaker standup, Granola’s transcript quality is unreliable: the published excerpt shows garbled phrasing and mistranscribed wording, and the report says the mishearing pattern recurs across early, middle, and late sections rather than being isolated to one moment.HappyScribe◐ MixedOn this ~25-minute multi-speaker standup, HappyScribe captured the vast majority of names, tools, and jargon correctly, and the report records only 1–2 misheard words.MeetGeek✓ WorkedIt transcribes a normal ~25-minute, ~10-active-speaker engineering standup mostly accurately, with only minor proper-noun/term drift noted in the report; one example given is "Madin" being misheard for "Mahreen".Otter.ai✓ WorkedOtter generated a full transcript for the standup and, per the report, captured names, tool names, jargon, and numbers correctly with minimal errors, making the transcript reliable for reference.
This evidence is published in
Real inputs and real outputs, no retouching · every cell queryable via API & MCP · aidemos.com