condensr turns long-form video into concise, searchable knowledge. Paste a video URL from YouTube, TikTok, Instagram, X, or any of the 12 supported platforms, and get a structured summary with key points, timestamped quotes, categories, and speaker identification. You can then ask follow-up questions in an AI chat grounded in the video's actual content.
I built condensr because I want to know too much - too many videos that I want all the info from. The announcements, the talks, the breakdowns all moved to video, and you can't skim a video. You either give up the evening or give up knowing. condensr fixes that trade: drop in the day's videos, get the breakdown, ask what you missed. Keeping up is now just a morning brief.
condensr runs its own GPU transcription engine with speaker diarization - it does not depend on platform captions, so it works on videos that have none. Summaries keep timestamped quotes that link back to the exact moment in the video, so every claim is checkable at the source.
I'm Joseph Decker, a site reliability engineer, and I have been my own first user since day one. Pricing is credit-based with one-time packs - no subscription. Details are on the home page, and the machine-readable summary lives at /llms.txt.
Questions? See Contact.