
Share
Krafton and Nvidia's collaboration ushers in a new era for gaming with an offline AI system for NPCs, enhancing realism and responsiveness in PUBG and inZoi without relying on cloud computing.
Krafton, the studio behind the popular battle royale game PUBG, has partnered with Nvidia to introduce a groundbreaking on-device AI system that will revolutionize NPC (Non-Playable Character) interactions. This new technology, built around a Small Language Model (SLM), is designed specifically for video games and will be integrated into both the PUBG franchise and the upcoming life simulator, inZoi.
The key innovation here is the development of an SLM that runs entirely on local devices, without any cloud dependency. This means:
The SLM is designed to handle a wide range of in-game scenarios:
PUBG Franchise:
inZoi:

Kang-Wook Lee, head of Krafton’s deep learning division, highlighted the potential impact of this technology during CES 2025. According to Lee, these advancements can help PUBG attract a new audience by adding layers of depth and engagement to an already mature game. For inZoi, the SLM promises to create a highly immersive and dynamic simulation environment.
Krafton demonstrated the capabilities of the new AI in a video showcasing PUBG gameplay. In one scenario, an NPC assists a human player by providing strategic advice and even performing dance moves. This level of interaction is unprecedented and highlights the potential for more engaging and realistic co-playable characters.
In inZoi, the SLM brings to life a world where NPCs have their own lives, making decisions based on real-time interactions and environmental factors. For example, a Zoi might decide to follow a new fashion trend or spread a rumor through the game's social network.
The collaboration between Krafton and Nvidia marks a significant step forward in AI-driven gaming. By leveraging local processing and advanced language models, these co-playable characters will offer players more immersive and dynamic experiences in both PUBG and inZoi. As this technology continues to evolve, we can expect even more innovative applications in the future of interactive entertainment.
Tags
Original Sources
About the author
Kai built ML infrastructure at a Bay Area startup before developing an obsession with transformer architectures and inference optimisation that eventually pulled him out of product work entirely. A stint at a compute research lab sharpened his instinct for what actually matters in a model release versus what is marketing. He writes from the inside — from the perspective of someone who has debugged the systems he is describing at three in the morning. He is allergic to hype and instinctively drawn to the unglamorous plumbing questions that everyone else skips over.
More from The Engineer →This Week's Edition
15 January 2025
133 articles
Related Articles

Smarter Engagement for Stronger Growth: How Payers Can Leverage AI to Do More with Less
Products & Applications · 3 min

Penn Medicine and K Health Deploy AI Clinical Agents to Enhance Patient Care
Products & Applications · 3 min

Wheel and b.well Partner to Build Turnkey AI-First Virtual Care Infrastructure
Products & Applications · 3 min
Related Articles

Smarter Engagement for Stronger Growth: How Payers Can Leverage AI to Do More with Less
Products & Applications · 3 min

Penn Medicine and K Health Deploy AI Clinical Agents to Enhance Patient Care
Products & Applications · 3 min

Wheel and b.well Partner to Build Turnkey AI-First Virtual Care Infrastructure
Products & Applications · 3 min
More Stories