Every AI Finder
All benchmarks
Protocol ready · controlled run pending

CLOVA Note vs Notta vs Fireflies: Meeting Test

Measure Korean transcription, speaker separation, decisions, and action items.

Common prompt

Transcribe the supplied 12-minute Korean product meeting with three speakers. Produce a verbatim transcript, a five-bullet summary, decisions, action items with owners and deadlines, and unresolved questions. Do not infer missing owners or dates.

Controlled conditions

  • Same WAV file with a human reference transcript.
  • No custom vocabulary on the first run.
  • Raw transcript and corrected transcript are both retained.

Published criteria

Korean transcript accuracy

35%

Counts omitted, substituted, and spacing errors against a reference transcript.

Speaker separation

20%

Checks whether three speakers and interruptions are assigned correctly.

Summary and actions

25%

Checks decisions, owners, deadlines, and unsupported additions.

Export workflow

20%

Measures processing time, correction time, formats, and collaboration.

Tool records

A controlled test has not been run yet. Raw outputs and scores are therefore hidden and do not affect tool ratings.

CLOVA Note (클로바노트)

Plan: Controlled test plan to be recorded

Protocol ready · controlled run pending

Tool details

Notta

Plan: Controlled test plan to be recorded

Protocol ready · controlled run pending

Tool details

Fireflies.ai

Plan: Controlled test plan to be recorded

Protocol ready · controlled run pending

Tool details