Are you tired of manually re-watching long webinars, tutorials, or meetings just to find where a new topic starts? Segmenter is a game-changing AutoHotkey tool that does the heavy lifting for you! By leveraging AssemblyAI for pinpoint-accurate transcription and the intelligence of OpenAI or Claude to analyze the text, Segmenter automatically detects natural topic shifts. It then splits your long-form video into perfectly crafted, bite-sized clips with descriptive titles., API calls, and modern GUI elements to build powerful AI desktop applications in AutoHotkey v2!

Updated on 2026/05/22 to Version: 1.3.0

AI Finds Your Best Clips 🤖 So You Never Scrub Footage Again

Stop Letting Your Best Content Rot Inside a 2-Hour Video

If you create video — webinars, livestreams, podcasts, training sessions, sales calls — you’re sitting on hours of footage you’ll never reuse. Not because the content isn’t good, but because manually scrubbing through a 2-hour recording to find, label, and cut the eight best teachable moments takes more time than the recording itself. Segmenter fixes that. Drop a video in, get back a list of timestamped, titled clips, click a button, walk away with separate files ready for upload.

The Problem

You hit “stop record” on a 90-minute training and the file lands in your library. You know there are five or six standalone teachable moments inside — but to actually pull them out you’d have to:

• Watch the recording end-to-end with a notepad open
• Write down timestamps and topics for each segment
• Load the file into a video editor
• Set in/out points for every clip
• Export each one with a meaningful filename

That’s a half-day of work per recording. Most creators just don’t do it. The result: thousands of dollars of evergreen content sitting on a hard drive, unwatched after the live event ends.

How Segmenter Works

Five steps. Most of them are watching a progress bar.

1. Drop your file. Drag a video (mp4, mkv, avi, mov, wmv) or audio (mp3, wav, ogg, flac) onto the Segmenter window.
2. Wait for transcription. Segmenter pulls the audio and ships it to AssemblyAI for a timestamped transcript. Cached locally — load the same file twice and you pay nothing the second time.
3. Tell the AI what you want. Pick a target clip length (Short / Medium / Deep-dive), an audience skill level, and a prompt template. Hit Go.
4. Review the segment list. A grid fills with start times, clip lengths, and AI-written clip titles. Double-click any row to preview in your player.
5. Cut. Select the rows you want, click Extract (fast, lossless) or Render (frame-accurate). FFmpeg cuts the clips on your machine — your original file never leaves your hard drive.

What Makes It Different

• Hours become minutes. A two-hour recording becomes a labeled clip list in the time it takes to transcribe it.
• You pick the AI. OpenAI (GPT) or Anthropic (Claude). Switch in Preferences whenever you want. No lock-in.
• Custom prompt templates. The default prompt looks for natural topic boundaries. But you can write and save your own — “find every FAQ moment,” “find every demo,” “find every objection” — and reuse them across hundreds of recordings.
• Six length presets. Any length, Short clips (under 1 min), Medium (1-5 min), Short-to-Medium (6-10 min), Deep-dives (10-20 min), and Long deep-dives (over 20 min). Pick the one that fits your output channel.
• Skill-level targeting. Tell the AI whether the audience is beginner, intermediate, or advanced. It changes what segments get picked.
• Two cut modes. Stream-copy for speed and zero quality loss, or full re-encode for frame-accurate cuts and stubborn codecs.
• Local extraction. FFmpeg runs on your PC. Only the audio transcript leaves your machine — your video file stays put.
• You own the API keys. No markup. You pay AssemblyAI and your AI provider directly, at cost.
• Dark theme. A clean dark-themed UI that doesn’t burn your retinas during a long editing session. Toggle to light in Preferences.

Who It’s For

If any of these sound like you, Segmenter pays for itself the first weekend you use it:

• Course creators chopping long webinars into modular lessons
• Podcasters pulling shareable audiograms and YouTube Shorts from long episodes
• YouTubers repurposing livestream archives into evergreen short-form content
• Educators & trainers extracting teaching moments from recorded sessions
• Sales & marketing teams turning customer calls and demos into highlight reels
• Researchers & journalists indexing long-form interviews

A Realistic Example

A 90-minute marketing webinar. You set “Target Clip Length: Medium (1-5 min),” “Target Skill Level: Beginner-friendly,” and the Default prompt. Click Go. Sixty seconds later you have a grid of twelve segments:

Select all rows, click Extract, walk away. When you come back you have twelve titled MP4 files ready for upload, podcast, course platform, or social.

Total work: about three minutes of clicking. Cost: under a dollar in API charges. Result: a week’s worth of social-media content from one recording.

Get Segmenter

Windows desktop app. Runs on AutoHotkey v2 or as a standalone .exe. Lifetime updates.

You bring your own AssemblyAI and OpenAI/Anthropic API keys — no monthly subscription, no markup, pay providers directly at cost.

The Bottom Line

If you record long-form video and only ever use the live cut, you’re throwing away the majority of the value you already paid to create. Segmenter turns those archives into shippable assets — one drop, one click, done.

Your back catalog is sitting there. The question is whether you want to keep ignoring it, or finally start mining it.