Comparative Test of Meeting Summary Generation: MeetingMinutes vs. DingTalk Flash Notes

Recently, I conducted a comparative test on the core meeting summary generation feature between MeetingMinutes and DingTalk Flash Notes.

For the test, I used a recording of a real department weekly meeting as material, lasting approximately 90 minutes with 5 speakers. The content covered typical meeting scenarios such as project progress updates, problem discussions, and next week's work arrangements. Both tools processed the same audio segment and were compared across three dimensions: generation speed, content quality, and structural completeness.

In terms of generation speed, DingTalk Flash Notes, supported by the DingTalk ecosystem, can quickly generate summaries after transcription is complete, with overall fast processing speed. MeetingMinutes also performs well in generation speed—producing a complete summary for a 90-minute audio in about 10 minutes. There is not much difference in speed between the two.

Regarding content quality, both tools can effectively extract the core content of meetings. MeetingMinutes's summaries are more precise in extracting key data and decision conclusions, automatically filtering out filler words and repetitive expressions, resulting in more concise transcripts. DingTalk Flash Notes' summaries are relatively comprehensive with more detailed records of the discussion process, but there is still room for improvement in the conciseness of key point extraction.

In terms of structural completeness, MeetingMinutes comes with over 50 summary templates, allowing users to select corresponding templates based on meeting types—such as project reviews, interviews, and retrospectives—to generate structured summaries with clear categorization and logical hierarchy. DingTalk Flash Notes' summary structure is relatively standardized, mainly organized by timeline and topics, with relatively fewer template options.