Roblox has announced a photorealism upscaler called Roblox Reality, which will generate high-fidelity visuals on top of the platform’s core engine.
In a blog post, Roblox vice president of engineering Anupam Singh explained that Roblox Reality integrates the existing Roblox Cloud and Game Engine with AI-driven Video World Models to enhance rendering details such as lighting, textures, and motion.
Singh said that Roblox Reality will “significantly reduce the development time, cost, and compute that is traditionally required for high-fidelity graphics.”
“This makes creating photorealistic games faster and more cost and compute efficient for our creators. Given the high compute cost, we realise there are challenges we need to solve before we can scale the Roblox Reality architecture.”
Singh added that Roblox is “working on solutions to optimise and improve efficiency for this architecture, enabling Roblox Reality to scale more affordably to millions of concurrent players.”
Roblox expects to release an early version of Roblox Reality later this year or early 2027.
Roblox CEO David Baszucki described the technology as a “hybrid architecture that marries the structured data and logic of the Roblox Engine and Roblox Cloud with the generative power of Video World Models.”
“We believe this will ultimately remove barriers to high-fidelity creation, allowing a team of three people to build a narrative-driven, photorealistic masterpiece in a single week.”
Roblox recently published its Q1 financials, reporting 39% revenue growth to $1.4 billion and a 43% increase in bookings to $1.7 billion.
The company also noted that recent mandatory age checks and child safety features reduced its 2026 growth outlook.
It now anticipates full-year bookings growth of 8% to 12% and revenue growth of 20% to 25%.
