UseShorts

Turning a one-hour podcast into a handful of vertical clips is normally an hour of scrubbing, cutting, reframing and captioning per clip. UseShorts compresses that into a short pipeline that runs on your own machine. You point it at a long video, it transcribes the audio (either from the existing YouTube transcript or via Deepgram with your own API key), and a cloud AI model analyses the transcript to score and rank the moments that would work as standalone shorts.

The distinctive part is the framing. Rather than exporting a fixed centre crop, UseShorts tracks a moving subject and keeps them composed in a 9:16 frame. It also handles the podcast case: it can start on the wide two-person setup for context and switch into Dual Crop, producing a clean close-up of each speaker as the conversation shifts. Similar smart reframing moves between A-roll and B-roll or widens the frame when the wider scene matters more than a tight shot. Captions are generated and can be edited and styled, and export supports several video layouts for different vertical placements.

It is a deliberately local-first tool for creators who don't want a monthly subscription. UseShorts itself is licensed once, and there is no monthly conversion cap: you bring your own OpenAI and Deepgram keys, so the AI usage is billed by those providers separately and at cost. The app includes brand profiles for saved caption styles, fonts, colours, watermarks, logos and end-of-clip messages, which makes producing consistent shorts for several clients far less repetitive. It runs on Windows 10/11 and Apple Silicon Macs (M1 or newer).