hckrnws

So I appreciate the aim here, but for me to trust any video editor, I need to see an example timeline that’s like, 30 minutes long with clips from at least 10 1080p video files and at least one effect on each track.

And for the record I wouldn’t consider that a stress test (a stress test would be more like 3 hours, 100 tracks, 4K and like a dozen precomps that are being reversed or something). That’s just to make sure this thing won’t fall over during casual usage.

You may be inclined to respond that your editor is targeting beginner editors, to which I’d note that beginner editors are MUCH less disciplined than experts when it comes to trimming footage, splitting things up into comps, pre-rendering chunks, using proxies, etc. Beginner editors (I’d know, I used to be one) will dump a 1 hour 4K-HDR iPhone video of a presenter speaking, and a screen recording of presentation slides they accidentally took in 4K60 into your timeline. Being able to demonstrate that you’ve got that level of memory management handled is what separates video editors people can use from mere “good ideas”.

Edit: Another thought, you call your product “Cursor for video editing”, and that’s a valid goal. But bear in mind that a LARGE part of why Cursor is successful is because they didn’t try to build an IDE from scratch. They got to absorb all of the nice UX (not to mention the extensive plugin ecosystem) of VS Code, and then spend their time on AI features. If that’s how you want to spend your time, you definitely don’t want to be building an editor from scratch.

I dont get this attitude. If there is some specific test you need before you'll trust it, what's stopping you doing that? I can understand your response if it was a paid product, but this doesn't even require you to install anything. Instead of appreciating the work that they've put in, you come right in with a negative comment that's not even based on trying it or knowing anything about it.

I’m on my phone at the moment, and didn’t want to judge this product on its iOS Safari performance, as that wouldn’t be fair.

With regard to my attitude: I’ve run into a lot of people who are trying to build “AI video editors”, and many of them don’t realize how intense a basic video editor actually is. It’s the kind of area where, the faster someone gets to the brick wall, the sooner they can start working on getting through it.

Furthermore I think it’s good for them to know that if achieving what I described seems daunting and they want to focus on the AI angle, it’s totally fine to fork an existing mature OSS video editor and just build the AI features on top. That’s what Cursor did for IDEs, and they’re finding a lot of success.

It's ok to wait til you've had a chance to try it before commenting.

They don't seem to me to be focusing on the AI angle at all. They mention one AI feature as coming soon.

Weird critique. Someone builds a quality pair of lightweight scissors, and you critique it like a boastful 5-axis CNC operator.

Like the rest of us needing to edit many couple of minutes long videos, there is massive gap in the market for something lighter weight. Look at Capcuts success over Adobe: 200+ million active user per month for Capcut.

Well, my very simple 3-minutes playing around with this were enjoyable.

Please add the ability to center images if you upload different-sized images, so the smaller images don't all clump in the upper left corner.

I'll try this with larger video clips later

sure :)

try with larger clips and let me know in discord / github discussions

Another remotion video editor with vibe-coded features? There are so many of them... A video editor in the long term needs to be mobile-first. A web video editor in the big year of 2025 is not going to move the needle. Capcut has a free desktop app that you must compete with. I think a better idea is building a mobile video editor, it's much harder to vibe code. To be the "Cursor for Video Editing," it's a must.

Capcut TOS change is very unfortunate. They now claim broad licenses over user content. (https://www.isabokelaw.com/blog/capcuts-new-terms-of-service...)

Also, what makes you think it is vibe coded? Is the app not functioning as you expected? Let me know and I'll fix them asap.

mobile-first it's how to degrade computing in first place.

How does it compare with OpenCut?

https://opencut.app

You should prepopulate the timeline with an in progress edit so people can jump in start playing with it.

How does it compare to https://omniclip.app/ ?

kimu offers a better user experience and is more intuitive. Also, we match the features with omni.

Our roadmap included automated captions, color grading and the like. We'll be a capcut alternative on the web.

Sounds great. CapCut is the editor I use the most.

>A cute AI agent (coming soon!)

Why??? You don't need this just because "AI" is popular right now, it will distract you from the goal of developing "video editor built for the web". It's really not going to improve the video editing experience.

our goal is to make a solid web video editor first and foremost. Then we'll try to make it super accessible.

I'm thinking of including AI features like captions, auto color grading etc. I get your point of forcing AI, we won't do that.

No AI for AI, but AI for making it more accessible and helping novice users to also make cinematic videos.

Captions would be a very useful feature and one of the top feature that paid platforms use as a hook for payment. There are enough models that can run client-side to make this good enough for social media captions, for example.

yeah. im working on a PR for it rn. If you like our work, join discord :)

So you are spending time shoe-horning "AI" into a web-based video editor, when you could have been creating PRs for actual video-editing functionality.

IMHO captions is not "shoe-horning AI" for video -- it's a critical requirement to be competitive with closed source editors and a great use case for local models.