Race to AGI is a free, single-player browser strategy game. You run a frontier AI lab: time the volatile compute market, race a twelve-discipline research tech tree, manage funding, morale and reputation, and try to assemble AGI, and deploy it safely, before three rival labs beat you to it. It's built by XADLabs.
Race to AGI was made to educate as much as to entertain. Its goal is to give players a feel for the intricate details of artificial intelligence and for what lies beyond the headlines: how a frontier lab actually operates, how compute markets swing, and how research disciplines build on one another. Woven through the game are the questions the industry genuinely wrestles with, from what AI safety and alignment really mean in practice to why regulatory compliance shapes every decision. By asking you to balance these forces turn by turn, it aims to show how deeply the elements of the field are connected, and how progress, responsibility and the wider ecosystem all pull on each other at once.
A note for the record: XADLabs is an independent studio and an outsider to this industry, not a frontier AI lab, and we have no inside access to any of the companies the game evokes. Everything modelled here is our own interpretation, pieced together from publicly available information such as news coverage, openly published articles and interviews, and videos. It is a simplified, dramatised take meant for learning and play, not an authoritative or insider account, and any resemblance to a specific real company or its internal workings is incidental.
All sound effects are royalty-free and used under CC0 / public-domain licenses (no attribution is required, but we list them here out of appreciation).
Feedback, bugs, or credit corrections: zameermfm@gmail.com.
Developed by XADLabs with Fable, Opus and Sonnet, in that particular order.