A free Descript alternative for turning long videos into clips
Descript is a full video and podcast editor that you edit like a document. Clips Kitty does one part of that job: it finds the best moments in a long video and turns them into captioned vertical clips, for free, on your own PC.
If you mostly use Descript to cut clips out of streams, podcasts or long videos, this page is for you. If you use the whole editor, the section on when Descript is the better choice is written for you too.
Clips Kitty is free & open source ❤️
It runs entirely on your PC with no fees. Please consider donating to help cover development costs and keep it free for everyone.
The short version
Switch for clipping if you turn long recordings into short clips regularly, you own a Windows PC with a graphics card, and you would rather not pay per seat or count media hours.
Keep Descript if you edit by transcript, record your screen or your podcast in it, share projects with a team, or work on a Mac. It does those well, and Clips Kitty does not do them at all.
Descript vs Clips Kitty, row by row
| Clips Kitty | Descript | |
|---|---|---|
| What it is | A clipper: finds the moments, crops to 9:16, captions them and writes the titles. | A full video and podcast editor you edit like a document, with AI tools. |
| Price | Free. AGPL-3.0, no paid tier. | Free plan; Hobbyist $16, Creator $24, Business $50 per person a month billed yearly ($24, $35 and $65 billed monthly); Enterprise on request. |
| Hours you can process | Unlimited. | Media hours by plan: 60 minutes on Free, then 10, 30 and 40 hours a month. |
| AI allowance | None to run out of: the AI runs on your PC. | AI credits by plan: 100 once on Free, then 400, 800 and 1,500 a month. |
| Watermark and resolution | Never a watermark. Exports at your source's quality. | Free exports 720p with a watermark; Hobbyist 1080p; Creator and Business 4K. |
| Clips from a long video | Automatic: it scores every moment on what was said, how the audience reacted and what is on screen, then crops and captions the best ones. | Create Clips from Hobbyist up, and the Underlord AI editor (limited on Hobbyist). |
| Editing by transcript | No. Its editor trims, cuts, mutes words, adds a hook, music and a watermark. | Yes. It is the heart of the product. |
| Translation and dubbing | 19 languages, free. | 30+ languages, on Business. |
| Where processing happens | Your PC, unless you choose cloud AI. | Their cloud. |
| Account required | No. | Yes. |
| Mac | No. Windows 10 and 11 only. | Yes, macOS 14 or later. |
| Hardware needed | 16 GB RAM, NVIDIA RTX 2060 or newer recommended. | 8 GB RAM; Windows 11 or macOS 14 or later. |
| Source code | Open. Auditable and forkable. | Closed. |
| The AI model | Your choice: local through Ollama, or a cloud model through OpenRouter on your own account. | Theirs. |
Descript plans, prices, media hours, AI credits, export limits and plan features checked on 1 October 2026 at descript.com/pricing, and the system requirements in Descript's help centre. Their plans change, and this table does not update itself. Only the rows above were verified: anything not listed here is something we did not check rather than something they lack.
What it actually costs over a year
Per-person monthly pricing hides the number that matters, so here it is written out.
- Hobbyist at $16 a month is $192 a year, Creator at $24 is $288, and Business at $50 is $600, per person, billed yearly. Billed monthly, the same three are $288, $420 and $780.
- Media hours decide the plan as much as features do. A weekly three-hour stream is about 13 hours a month, which is past Hobbyist's 10 and inside Creator's 30.
- On your own hardware the marginal cost is electricity. A mid range graphics card draws a couple of hundred watts under load. That is pennies per stream, whatever the length.
- Cloud AI is pay as you go, paid to OpenRouter directly. In our test, clipping a 2-hour stream with Gemma 4 26B-A4B for the AI and Whisper large-v3-turbo for the transcription, both through OpenRouter, cost about 11 cents. What you pay depends on the models you pick and the length of the video.
- The crossover is what you use it for. If you need Descript's editor anyway, its clip tools come with it. If clipping is the whole job, you are paying for an editor you do not open.
Where your video goes
A cloud editor needs your media on its servers before it can work on it. That is how the architecture works, not a criticism of any particular company.
It matters more in some situations than others: unreleased footage, sponsor reads under embargo, a guest who signed something, a client's rushes. Clips Kitty reads the file where it already sits, and by default nothing leaves your machine. The full detail is in the privacy policy, which is short and specific.
Two exceptions, both off until you switch them on. If you choose cloud AI, the transcript, and the audio when transcription runs in the cloud, go to the service you signed in to; the video itself stays on your PC. And publishing a finished clip sends that clip to the platforms you post it to.
Cloud AI, on your own account
Local AI is the default, and it needs a reasonably modern graphics card. When a PC is too old for that, Clips Kitty can run the clip scoring and the transcription in the cloud instead.
Sign in with your OpenRouter account and pick the models; it is billed to your own account, with no Clips Kitty server in between. Everything else, the cropping, the speaker tracking, the captions and the rendering, still happens on your PC. The detail is in the AI backends guide.
Descript's AI runs through OpenRouter too: Descript is one of the top apps on OpenRouter's public app rankings. The difference is who pays for it. With Descript, the AI is paid for inside its plans and AI credits. With Clips Kitty you pay OpenRouter directly, at OpenRouter's own prices, and Clips Kitty takes nothing. In our test, a 2-hour stream cost about 11 cents.
When Descript is the better choice
This section is here because it is true, and because you will find out anyway.
- You edit by transcript. Deleting words to cut the video is Descript's core idea, and Clips Kitty has nothing like it.
- You record in it. Screen recordings and podcast recording are part of Descript. Clips Kitty starts from a finished video.
- You work as a team. Clips Kitty has no shared projects: clips live on the machine that made them.
- You use a Mac. Clips Kitty is Windows only, and no future release fixes that soon.
- You do not have a graphics card and do not want cloud AI. It runs on the processor, but slowly enough that most people give up on it.
- You want support with someone behind it. This is alpha software, the installer is unsigned so Windows shows a warning the first time, and the support channel is a GitHub issue tracker.
What Clips Kitty does that is not on that list
- Streams, not just recordings. Paste a Twitch or Kick VOD or a YouTube link, or watch a channel and let it clip each new video by itself.
- Game streams and reaction videos. The streamer's webcam and the game together, in eleven layouts.
- Soccer matches. Goals, saves and cards as clips, with a match recap and team or player reels.
- It learns a creator. Running jokes, catchphrases and ongoing storylines feed into how their moments are scored, and it stays on your PC.
Questions people ask before switching
Is there a free alternative to Descript?
For turning long videos into clips, Clips Kitty is free with no paid tier at all. It runs on your own Windows PC, so there are no media hours to count, no watermark and no per-seat price. It is not a replacement for Descript's whole editor: it does the clipping part of the job, not editing by transcript.
Is there an open-source alternative to Descript?
Clips Kitty is open source under the AGPL-3.0 licence. You can read every line, including every line that touches the network, change how clips are scored, or fork it. Because the licence cannot be revoked, a copy you have today keeps working whatever happens to the project.
Can Clips Kitty replace Descript?
If you use Descript mainly to cut clips out of streams, podcasts or long videos, yes. If you edit by transcript, record your screen or podcast in it, or share projects with a team, no: Clips Kitty does none of those, and Descript does them well.
Does Clips Kitty upload my video anywhere?
No. Transcription, scoring, speaker tracking, captions and rendering run on your own machine by default. If you choose cloud AI, the transcript, and the audio when transcription runs in the cloud, go to the service you signed in to; the video itself stays on your PC. Publishing is optional and off by default.
Does Clips Kitty work with OpenRouter?
Yes, as an option. When your PC is too old to run the AI itself, sign in with your OpenRouter account and the clip scoring and transcription run on models you pick. OpenRouter bills your own account directly and Clips Kitty takes nothing: in our test, a 2-hour stream cost about 11 cents. Local AI stays the default.
Does Descript use OpenRouter?
Descript is one of the top apps on OpenRouter's public app rankings, so its AI runs at least partly through OpenRouter. With Descript, that AI is paid for inside its plans and AI credits. With Clips Kitty's cloud AI you pay OpenRouter directly, at its own prices.
Do I need a graphics card to run it?
It runs without one, on the processor, but slowly enough that most people will not enjoy it, unless the AI runs in the cloud through OpenRouter. An NVIDIA card makes the largest single difference. The published minimum is Windows 10 or 11, 16 GB of RAM and an RTX 2060 or newer, and the 16 GB is a requirement rather than a suggestion.
Try it on one long video
The honest test is your own footage. Run one stream or episode through it and see whether the moments it picks are the ones you would have picked.
Descript is a trademark of its owner. Clips Kitty is an independent open-source project and is not affiliated with, endorsed by, or connected to Descript. The comparison above is offered as fact, with the date it was checked, so you can verify it yourself.