nowfound

Work · August 15, 2026

Drift Cut

Drop in a 3D model and a song. Get a clip cut to the kicks.

What it does

Drop in a 3D model and a song. Drift Cut finds the kicks offline, cuts a camera edit against the song's arrangement, and encodes it vertical to MP4 in the browser. A template tool cuts footage that already exists and a generative tool makes pixels from a sentence. This films the object you hand it, and nothing is uploaded. No model? The Music visualizer takes a track and its sleeve instead. Free today: eight worlds, 1080p at 60, three clips a day. Needs a desktop browser.

Drop in a 3D model and a track. Drift Cut cuts a camera edit to the beat and exports it vertical, entirely in your browser.

Drift Cut turns a 3D model and a song into a vertical video, cut to the beat. Drop in a model and a track: the kicks are found before a single frame is drawn, so a cut lands on the transient rather than near it. The camera works hard over a drop and holds back over the quiet, and the MP4 is encoded in your own tab at up to 4K. Four bars at 124 bpm. Twenty-one kicks found in the track itself, and six cuts, each one of them on a kick rather than on a grid. Twenty-six of them, and every poster below is a frame of that world photographed on a real GPU rather than an illustration of it. One of the twenty-six is a layout rather than a place, and it is the one that needs nothing modelled. A song…from driftcut.app

Does the same job

all alternatives →
  • RenderrApr 2026 · ▲86

    Automatically cut, structure, caption + adds music vids

  • Beatwave2023 · ▲111

    Create custom audio visualizers in just a few clicks

  • MV
    Music video using AI (Stable Diffusion) and a text editor – You can too2022 · old.reddit.com · ▲9

    I went nuts on Sunday after stumbling on this animator script for the automatic branch: https://github.com/Animator-Anon/Animator And then it's a ton of keyframing. I also found that the timings all have to be earlier than the timestamp you want to account for how long it takes to denoise to your new scene - so like .4 gives smoother but longer, .7 is snappy. To get stuff to keep showing up and stay dynamic you can zoom out continuously, it needs a UI, badly. If you're on windows you'll need ffmpeg too. I watched this when it came out last week:…

  • DeadMod Studio Jul 2026 · deadmod.com · ▲8

    Make a music video before your song ends

  • Supercut - AI Video Editor22d ago · supercut.cc · ▲12

    Type the edit. Record your screen. Zero uploads.

  • Music2storyboardgen5d ago · music2storyboardgen.com · ▲2

    Any track → an AI music video. Free local GPU included.

More work this month

the category →
  • Let agents source clips from terabytes of your local video

    Work · 18d ago · clipto.com

  • Free local transcription that is 100% Private

    Work · 17d ago · hynote.ai

  • The app store for voice native apps that lives in your notch

    Work · 28d ago · voiceos.com

  • Ask any question, get a video back instantly

    Work · 24d ago · scrimba.com

  • Handle all of the work before, during, and after meetings

    Work · 17d ago · calendly.com

Launched alongside, August 2026

the whole month →
  • TL

    Life & fun · 10d ago · louisabraham.github.io

  • Hey Noah641

    A proactive AI executive assistant for founders

    AI · Aug 2026 · heynoah.io

  • Let agents source clips from terabytes of your local video

    Work · 18d ago · clipto.com

  • SA

    Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…

    Life & fun · Aug 2026 · toneyalexander.github.io

  • AdAnt AI608

    Claude for viral, high-converting social ads

    AI · Aug 2026 · adant.ai

  • I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.

    AI · 17d ago · simedw.com