Ryan Craig

Data Eng

[Ryan Craig @RyyanCraig]

Unimaginably proud to be part of this incredible team. Real talent, real engineers, spearheading what we firmly believe to be the next frontier of entertainment all while having it run at 60fps on an RTX3060. More good things to come keep your eyes peeled👀

[Overworld @overworld_ai]

While many frontier world models require large-scale GPU infrastructure, Waypoint-1.5 is the only world model designed to run on consumer devices.

The paper includes benchmarks across a wide range of RTX GPUs, demonstrating real-time performance on hardware people already own.

[Overworld @overworld_ai]

While many frontier world models require large-scale GPU infrastructure, Waypoint-1.5 is the only world model designed to run on consumer devices.

The paper includes benchmarks across a wide range of RTX GPUs, demonstrating real-time performance on hardware people already own.

[Louis Castricato @ lovecraftian horrors @lcastricato]

This paper has been cooking a LONG time, we're incredibly excited to finally share some of our research findings. A *lot* more to come in the coming months. Upwards and onwards.

[Shahbuland @shah_bu_land]

Not many can say they have a video diffusion model running at 60fps, much less on an RTX 3060 😜

[Overworld @overworld_ai]

Through a custom distillation approach built on self-forcing, we reduced generation from 60 denoising steps to 4 steps with minimal quality degradation, enabling longer stable rollouts.

Learn more:
📄 Find the paper here: https://t.co/fgX2KbXKZt
💻 Code: https://t.co/wXsPdSZ6Rq
🤗 Model: https://t.co/7WL2pXfFnc

[Overworld @overworld_ai]

The Waypoint-1.5 technical paper is now live.

Waypoint-1.5 is a real-time video diffusion world model designed to run on consumer GPUs, bringing interactive world models closer to practical, accessible deployment.

[Overworld @overworld_ai]

Our CEO, @lcastricato, recently sat down with the @AP to discuss how Overworld came together, where world models are headed, and what we're learning as this category continues to take shape.

Take a look: https://t.co/qv0tOOg1j9

[anm_ol @anm5704]

Thrilled to announce Waypoint-1.5, the next step towards diffusion world models running locally. Working on this release was so much fun.

onto the next one now :)

[Overworld @overworld_ai]

Today we’re releasing Waypoint-1.5.

An update to our real-time diffusion world model designed to run interactively on consumer hardware.

[Overworld @overworld_ai]

The hardest part of shipping a world model isn’t the model.

It’s the gap between what users type and what the model expects.

It’s not just text → output.

[Overworld @overworld_ai]

We’ll be in SF next month for GDC and GTC, and we are bringing a new demo with us 👀

If you care about world models, real-time systems, or the weird edge cases that only show up when you try to run this stuff live, let’s grab time while we’re in town.

We want to see what you come up with

[Teknium 🪽 @Teknium]

My ol' boss @lcastricato released a world model 👀👀

[AK @_akhaliq]

Waypoint-1

Real-time Interactive Video Diffusion from Overworld

[Machine Learning Street Talk @MLStreetTalk]

Today @overworld_ai released a research preview of their real-time, local-first world model built for interactive, playable AI-worlds 60fps, locally run, all on consumer-grade hardware with @shah_bu_land (Co-founder & Head of Research) and Andrew Lapp (Member of Technical Staff).

[Ryan Craig @RyyanCraig]

@overworld_ai is out here creating THE best interactive world model… that runs consumer hardware (yes seriously)… and is open source. Genuinely what more could you what?

[Overworld @overworld_ai]

Today, we’re releasing a research preview of our real-time, local-first world model built for interactive, playable AI-worlds

60fps, locally run, all on consumer-grade hardware.

Come take a look ⬇️

[Shahbuland @shah_bu_land]

I have spent years building AI systems, yet I still felt that something was missing.
Watching outputs appear is not the same as being inside a world that responds to you.
This is the first step toward closing that gap.

[∿ @somewheresy]

This is phenomenal. When these models mature it will open up an entirely new lane of data for RL

[Jimmy Apples 🍎/acc @apples_jimmy]

Open source, interactive world model

Soon I guess we can prompt to play as a researcher in an ai lab that endures constant drama

[Alberto @taiuti]

The latency here is great. Check it out.

World Models are here

[Konsti Wohlwend @konstiwohlwend]

you can just prompt your own games now?