A two-hour recording should not take two hours to clip. DecaTrend processes the full timeline, finds ranked moments, and returns a ready-to-review clip list — typically around 20 minutes of processing time for a 120-minute source file.
Best for educators, course creators, webinar hosts, conference recappers, and long-form interviewers who sit on hours of unused footage.
Scrubbing a two-hour file to find three clips takes longer than recording the content itself. Editors often stop at the first few obvious moments and miss stronger material buried in the middle or end. DecaTrend scans the entire file so no segment goes unreviewed.
After upload, DecaTrend transcribes the recording using faster-whisper, segments the transcript into candidate clips, scores each on six signals — hook strength, retention potential, emotional arc, caption density, audio energy, and trend fit — and returns a ranked list. A 90-minute webinar typically produces between 8 and 15 scoreable clip candidates.
A standard two-hour recording processed through DecaTrend returns approximately 12 ranked clips in about 20 minutes of machine time, versus 3 to 4 hours of manual review and editing for the same output. The ranking means you can decide what to post in a single pass instead of watching every candidate clip in full.
Every clip includes auto-generated captions from the same faster-whisper transcription pass, exported as VTT. Clips export as vertical 9:16 MP4 files ready for TikTok, Reels, and Shorts. Pro and Studio plans also export a portable project.json timeline that opens in OpenCut for frame-accurate trims.
DecaTrend supports uploads up to 5 GB. Most webinar and lecture recordings fall well within that limit. Very long livestream VODs may need to be split before upload.
A 60 to 120 minute recording usually produces 8 to 15 ranked clip candidates. The exact count depends on how densely packed the narrative is — interview-style content with multiple speakers tends to surface more distinct moments than solo lecture recordings.
Yes. DecaTrend returns the full ranked list and you select which clips to export or download. You are never locked into taking all generated clips.
Yes. DecaTrend processes the audio and video track together. Segments where the presenter speaks directly to the camera with strong delivery tend to score higher than screen-share-only segments.
DecaTrend is an AI clip maker for long videos, podcasts, streams, and interviews. Turn raw recordings into ranked clips with captions, trend matching, and editable exports.
DecaTrend transcribes podcast recordings with faster-whisper, detects speaker turns and quotable moments, and returns ranked short-form clips with editable captions.
Turn sales calls, coaching sessions, and webinar recordings from Zoom or Google Meet into short social clips with DecaTrend. Includes a privacy note on participant consent.