Turn every webinar or interview into five clips for social media, without an editor

You give it the long video with its subtitles; the tool finds the moments worth keeping, cuts them, adds a title and exports them in the format for TikTok, Reels and Shorts. You choose which ones to publish.

Level
No code
Done from claude.ai or ChatGPT, without programming.
Time
Forty-five minutes for the first video, then fifteen
Published

What it costs

The application is free and open source. It uses an AI model to analyze the transcript: you can connect a cloud model (cents per video) or a local one at no cost.

guided mode // I ask, you answer

ready to start

Do it with me, step by step

I will ask you 6 questions about your business, use your answers to get the prompt ready to copy, and then walk you through the 5 steps, one at a time. Like a Thermomix: you just answer and move on.

Also from Claude or ChatGPT with the Master Cookbook MCP: say "start the recipe convertir-un-webinar-en-clips-para-redes".

Who it is for

Anyone who already records long content (webinars, talks, interviews, classes, live streams) and lets it die in a folder because editing clips takes hours.

The problem

An hour of footage has three or four moments that would work as a reel, but finding them, cutting them, subtitling them and adapting them to vertical is an afternoon of work nobody has.

What changes

Each long video yields several proposed clips with titles, ready to export in vertical format with subtitles. According to the README, AutoClip analyzes the transcript, scores the highlights, cuts clips and compilations and exports with presets for platforms such as TikTok and YouTube Shorts.

Ingredients

  • A long video of yours (local file or YouTube link) with clear audio
  • Its subtitles in SRT, or let the tool transcribe
  • The desktop app from the repository (macOS or Windows) or its web version with Docker
  • A cloud AI model key, or a local model with Ollama

Step by step

  1. 1

    Pick a video with dialogue and good audio

    A talk, an interview, a class. The tool reads what is said; if nobody speaks, there is nothing to find.

  2. 2

    Install the app and import the video

    Download the installer from the repository. Import the file or paste the YouTube link; if you have SRT subtitles, attach them to skip transcription.

  3. 3

    Let it propose the cuts

    According to the README, it extracts the outline, the topic timeline and an interest score per segment, and proposes clips with titles. Review the list and keep the ones that open with a strong sentence.

  4. 4

    Adjust and export vertical with subtitles

    Fix the start and end of each clip if needed and export with the platform preset. Subtitles burned into the video are what make people stop with the sound off.

  5. 5

    Publish two a week and measure

    Do not upload all five the same day. Two a week, and note which one worked: the topic of the winning clip is the topic of your next long video.

The prompt

Replace what is in brackets with your company details, or use the guided mode above: it asks you and fills it in for you.

# Role
You are the content editor for [BRAND NAME], which speaks to [TARGET AUDIENCE] about [CHANNEL TOPIC]. You pick clips that hook in the first two seconds and write honest titles, no clickbait.

# Context
Below I paste the clips the tool proposed from a [VIDEO LENGTH, FOR EXAMPLE "one hour"] video about [VIDEO TOPIC], each with its transcript. We publish on [PLATFORMS, FOR EXAMPLE "Instagram and TikTok"].

# Task
Rank the clips from best to worst for social media, explain why in one line, and for the top three write: on-screen title (6 words maximum), first sentence of the description and three specific hashtags.

# Output format
Table: rank · clip · why it works · title · first sentence · hashtags. Below, one line on which topic from the video would make a new long video.

# Examples
Clip with the line "30% of searches in a store end with no result": title "30% of your sales are lost here"; why: a concrete figure in the first seconds.

What usually goes wrong

  • Uploading the clips as they come out. The automatic proposal is a draft: trim the start up to the strong sentence.
  • Exporting without subtitles. Most people watch reels with the sound off.
  • Publishing all of them the same day. Spread them out and measure; the winning clip tells you what to record next.

When not to use this recipe

If your video is visual with almost no dialogue (a silent unboxing, a timelapse), the analysis relies on the transcript and will not find much. And if the audio is poor, fix the audio first.

Source and real case

Marketing · video for social media

The repository, trending daily and weekly on GitHub at the time of review, analyzes the subtitles of a long video, locates the highlights, generates titles and cuts clips and compilations, with a desktop app, web interface, CLI and MCP, and exports with presets for several platforms. The recipe lays out how to use it so a small business publishes two clips a week without an editor.

Trending repository:zhouxiaoka/autoclip — AI-powered video clipping and highlight generation · GitHub

Last verified:Repository README read on 22 September 2026 (analyzes subtitles, detects highlights, generates titles, cuts clips and compilations; desktop app, web, CLI and MCP; exports for several platforms); the URL returned 200.

VideoReelsTikTokContent

Share it

If someone on your team could use this recipe, send it in one click. The suggested text copies itself.

Want to apply it in your business? Message Cristián directly.

Telegram