Getting started

Install Podcli, drop in an episode, and get captioned clips, a content package, and performance analytics. A guided walkthrough with real screenshots.

Podcli runs on your laptop. Drop in a podcast episode and it transcribes, finds the moments worth clipping, crops to the speaker, and burns captions in. Nothing leaves your machine.

This walkthrough covers the whole flow, start to finish.

Install#

No prerequisites. The installer fetches a self-contained binary, and the first run provisions everything it needs (Python, Node, FFmpeg, whisper.cpp, models) into a managed folder.

macOS and Linux

Windows (PowerShell)

Then run podcli and choose Open Web UI. The studio opens at http://localhost:3847.

Prefer the terminal? Everything here also works from the CLI.

1. Drop in your episode#

Open the studio and go to New episode. Drag in a local video, or paste a YouTube or direct link and Podcli downloads it. Pick a transcript engine (Whisper, or paste your own), then set caption style, crop, and format. The phone preview updates live.

2. Get clips automatically#

Podcli scores the transcript against your knowledge base, finds the moments worth clipping, and cuts filler into multi-segment clips. Every clip comes out captioned, framed on the speaker, and on-brand. Review them in the Library, grouped by episode.

3. Generate the content package#

Open any clip for a publish-ready package: title options, a description, tags, hashtags, and a thumbnail you can regenerate or reframe. Copy what you need, or change the caption style and re-render.

4. Track what actually works#

Link clips to YouTube and sync views, retention, and CTR back into the studio. Analytics shows what holds viewers by content type, caption style, and length, so the next batch is better than the last.

Where to next#