A personal studio
for shorter videos.
YTClip helps Ben find relevant source videos, choose excerpts, add captions and manage uploads to an authorised YouTube channel.
It runs on the operator's own server, with local speech transcription and a local language model. Source filters, clip lengths, caption themes and publishing schedules are controlled through a private studio.
Find relevant videos
Watch a selected channel or search a topic. Apply upload-date, duration and optional popularity filters to choose sources for processing.
Prepare excerpts
Transcribe speech, propose multiple excerpts and produce portrait clips with fixed captions that highlight the current word. Local checks inspect the rendered audio and picture.
Manage uploads
With the account owner's permission and publishing enabled, upload clips privately first, then manage their public visibility and schedule through YouTube API Services.
One operator. A private workspace.
YTClip is a personal project operated by Ben, @BenjaminIannotta. It currently serves one authorised account and has no public registration. The project is being configured and tested before unattended publishing is enabled.
Permission stays with the owner
Google handles account sign-in and consent. YTClip uses the granted YouTube permission to check the destination channel, upload clips and manage their visibility. The owner can pause processing, disable publishing or revoke Google's authorisation.
Only media the operator is entitled to use should be processed or published. Finding a video does not grant permission to reuse it.
Local AI, clear limits
Speech transcription and clip selection run on existing models hosted on the operator's server. This is inference: using a model to produce a result. YTClip does not train or fine-tune models on account data or source content.
Captions and suggested excerpts can contain mistakes. Automated checks do not establish copyright permission, factual accuracy or audience appeal.