In our research lab, we are building “real-time dreaming” - the ability to generate fully playable video worlds prompted from any text or image. Our real-time, action conditioned world model (currently running internally at 16fps at 832x480p) is trained on a combination of data, including proprietary Roblox 3D avatar/world interaction data. World models are different from multiplayer engines in that they store state and memory in video latents. Roblox is multiplayer, and we are actively researching optimal ways to simultaneously store state for thousands of players, and keep them in sync with their environment. Our world model leverages database technology which stores all user interactions on Roblox in a vector format that can be used to re-render video and interaction from any camera angle. We see several immediate uses for our Roblox world model. We will use it side-by-side text, image and video prompts as a way to launch auto-generation of immersive worlds. In Roblox Studio, a creator could walk around and use prompts to “paint” a world and then convert it into a 3D representation or direct to Roblox native as a way for many people to play simultaneously. All of this comes alive as we explore the notion of a “Dream Theater” - where one user is dreaming, while others watch and prompt them. 2/4

We are delivering on our promise and releasing Hytale into Early Access on January 13, 2026, with all the consequences that come with that decision. In just a few weeks, we merged roughly 300 branches, modernized a four-year-old code & asset base, scaled the team from 8 people to 50 core developers plus ~250 helping hands, tracked close to 1,000 bugs and fixed hundreds of them, added Mac and Linux support, and built an Exploration Mode that simply did not exist before. When we started the mission to save Hytale, we made a clear promise: Launch without any delays, at any cost. We are keeping that promise, even though it is a herculean task. In an alternate timeline, the most intelligent production decision would be to delay launch by 3-4 months, complete the obvious work in front of us, and avoid the overhead and distractions of maintaining a live product. The game would objectively be better on day one. That is not what we are doing. Moving this fast also comes with very real downsides, and you have already seen them. We make visible progress every week, sometimes so much that key community members work overtime to keep up with the news - but you have also seen us drop the ball. Some concrete examples: 1- Name reservations were intended to reward longtime fans. Still, we lacked sufficient time and information to verify the email list's integrity, and some longtime fans were missing from it. 2- The username generator produced unintended results, and our name filters failed to catch some unacceptable cases. 3- The payment system continues to report issues. 4- Communication regarding avatar customization, Steam release, modding, and pricing was insufficiently detailed, leading to misinterpretation or confusion. 5- Blog posts and technical write-ups have landed on an unpredictable schedule. 6- Game server providers, server owners, creators, and business inquiries are not receiving responses as quickly as we would like or as they deserve. This is the tradeoff we all have to accept: You can’t make an omelette without breaking a few eggs. We wish we had more time to double-check every step and spend more time on communication, but that time comes at a steep opportunity cost. Extra polishing reduces the overall impression of Hytale on January 13. Our primary focus is the game. The team is in a constant release cycle. Every week until launch is tightly scheduled, and we are doing this with a core team of around 50 people assembled in weeks, not a massive AAA organization. That means we are taking hundreds of calculated risks every day. Most pay off. A small fraction doesn’t - and when they don’t, the impact is very evident. If I could ask for one thing from the community, it’s this: Please keep reminding each other what this pace implies - especially newer players who were not around for the #SaveHytale moment and may not know what the Early Access experience is meant to be for us. Some people will not be able to get their desired username. Some will run into payment issues. Some information will arrive later than it ideally should. This is not because we don’t know better; it’s because we are choosing to keep a promise. For us, Early Access success looks like this: you start a world, and the first impression and immersion leave you thinking, “This is much better than they said.” After more time, you also start to see the rough edges - and still end your first day thinking, “This game has a lot of potential. I’ll come back and keep up with updates,” or get you inspired to get into making maps, content, mods or a minigame server. As you will be playing through the first week of Hytale, we will already follow up with the first patch. We will release updates rapidly in the first few weeks, with pre-releases available more frequently. You start on the 4-year-old archeology project, but within months, you will find yourself in a flow of rapid development, previously unthinkable. We appreciate that the community has high expectations for Hytale, but please remember that we are fulfilling a promise we made early on: to release as soon as possible. That is what the core community asked for. Our biggest fear right now is not criticism - it’s that the story of #SaveHytale gets forgotten during Early Access, and the team ends up wishing we had simply delayed those extra four months instead of keeping our word.

Hytale Early Access JANUARY 13, 2026