Studio-quality computer graphics made easy
Hey, y’all! Tony here from Dream3D– excited to share what Justin and I have been working on!
Dream3D is an AI-powered 3D design tool that makes it easy to create studio-quality computer graphics.
Creating 3D renders is an expensive and time-consuming process that can take teams of highly trained specialists months, if not years, to produce a final result.
In current rendering methods, input quality = output quality. A designer or artist needs to spend days building a scene, component by component, to get a final rendered result.
This leads to
Dream3D replaces complex workflows with text-driven commands and intuitive interfaces powered by our custom AI models.
Unlike existing generative tools, we give users explicit control over their scene, unlocking professional use-cases.
We let designers quickly build out 3D scenes by describing them in plain English– “a jetski in front of a building.“ We then let them transform that scene into a high-quality image with just another description of what they want– “photorealistic yellow jetski, evening sun, 80s magazine ad”
Here’s an example of transforming a simple car scene into production-ready concept art:
Or a full landscape from just a couple of boxes:
This means you can quickly iterate through multiple concepts and styles, maintaining the same scene composition, without having to spend time & energy building each scene from scratch.
It's fast and fluid so you can focus on what you want to create.
Right now, we let designers create static renders. But soon, we’ll be adding support for animation and video export.
In the longer term, Dream3D will be a complete 3D creative suite– letting you create full games, movies, and simulated worlds.
Justin and I are both second-time YC founders.
Justin’s last company, Bayes (S19), was a design-focused data analytics tool acquired by Airtable. He has a degree in Computer Science from Columbia.
Prior to Dream3D, I led the development of the machine learning platform at Clubhouse and my previous company, Inokyo (S18), built computer vision to automate retail checkout.
I first got interested in computer graphics and storytelling while I was studying at Carnegie Mellon. But, I found my creativity stifled by the complexity of the tools.
Now, we’re combining our experiences building prosumer tools and bringing machine learning research to production to make it easy for anyone to create studio-level computer graphics and build the tool I always wish I had.
If you do 3D design work (or want to), sign up here!
Or if you’re interested in researching hard problems across computer graphics and AI, and/or building delightful tools, send me an email at firstname.lastname@example.org