What’s New for Game Developers in NVIDIA RTX: DLSS 4.5 for UE5 and Multilingual AI Characters
Mirrored from NVIDIA Developer Blog for archival readability. Support the source by reading on the original site.
What’s New for Game Developers in NVIDIA RTX: DLSS 4.5 for UE5 and Multilingual AI Characters
NVIDIA RTX provides game developers with direct paths to AI-driven characters, frame generation, and ray-traced rendering. This post walks through a meaningful set of recent updates across the RTX ecosystem. NVIDIA ACE expands its multilingual AI character capabilities, making it easier to ship conversational NPCs. NVIDIA DLSS 4.5 arrives as an Unreal Engine (UE) plugin, bringing Dynamic Multi Frame Generation, a new 6x mode, and the second-generation transformer model for Super Resolution. And the NVIDIA RTX Branch of Unreal Engine (NvRTX) gets a stability-focused 5.7.4 release that tightens compatibility across NVIDIA RTX technologies.
NVIDIA ACE now supports multilingual, autonomous game characters
Traditional scripted dialogue and rigid NPC behaviors are no longer state-of-the-art when it comes to engaging in immersive interactions in open worlds. When characters lack reactivity or fail to acknowledge the context of a scene, immersion breaks. To address this, NVIDIA ACE provides a suite of technologies that transform static NPCs into dynamic, conversational agents—all powered by RTX GPUs.
ACE now features new multilingual models to ensure a global audience can experience a next level of immersion with game characters in their native languages. Access the latest models optimized for on-device performance with the NVIDIA In-Game Inferencing (NVIGI) SDK 1.6 update.
- Qwen 3.5 4B: This cutting-edge small language model (SLM) delivers low latency, contextually aware responses across 201 languages and dialects, ensuring characters react to player instructions with best-in-class accuracy.
- NVIDIA Riva Parakeet TDT 600M ASR: This high-performance speech recognition model enables characters to understand and transcribe 25 languages.
- Chatterbox Multilingual 500M: Supporting 24 languages, this model provides characters with expressive, emotionally resonant voices that pulls a global audience deeper into their gaming interactions.
In addition, NVIGI 1.6 enables you to quickly prototype Day 1 models with new functionality that connects to a locally running llama.cpp server.
Ready to build? Download NVIGI to integrate these technologies directly into your title.
NVIDIA DLSS 4.5 Unreal Engine plugin adds support for Dynamic Multi Frame Generation
The new DLSS 4.5 Unreal Engine plugin now supports DLSS 4.5 Dynamic Multi Frame Generation, Multi Frame Generation 6x Mode, and the second-generation transformer for Super Resolution.
The plugin provides direct access to the latest Super Resolution and Frame Generation updates. Built on Streamline, the plugin offers a consistent integration path across DLSS features, allowing developers to selectively adopt capabilities like Ray Reconstruction or Dynamic Multi Frame Generation. Updated APIs, documentation, and sample code help reduce integration time and make it easier to bring DLSS into both new and existing projects.
Updates to NVIDIA RTX Branch of Unreal Engine 5
Version 5.7.4 of NvRTX is now available to download. This release updates NvRTX to the latest Unreal Engine 5.7.4, and improves stability and compatibility for developers integrating RTX features into UE5 projects.
The NvRTX 5.7.4 update includes:
- RTX Mega Geometry shader compile fix: Addressed a shader compilation issue affecting RTX Mega Geometry on non-DX12 platforms
- Opacity Micro-Map (OMM) fixes: Stability and correctness improvements to OMM, which accelerate ray-traced alpha-tested geometry such as foliage and vegetation
- Substrate material improvements: Compatibility updates between Unreal Engine 5 Substrate material framework and NvRTX rendering paths
- NvAPI fixes: Resolved issues in the NvAPI integration that exposes NVIDIA driver-level features to Unreal Engine
- Code-base rebase: NvRTX brought up to Unreal Engine 5.7.4
- Documentation updates: Refreshed integration and migration documentation for the 5.7.4 release
Learn more
What if NPCs stopped following dialogue trees and started performing? Find out at Unreal Fest 2026, where NVIDIA will showcase a new workflow for building autonomous game characters. The workflow separates the role of an NPC from its performer to create deeply reactive, always-in-character AI personalities. Join us on June 17 for the session Ready, Set, Action: Why Your Next NPC Should Be Their Own Method Actor.
Using synthetic training data generated by large teacher models, developers can distill character behaviors into lightweight, deployable LoRA-powered performances that run locally on RTX hardware. The demo features a fully interactive MetaHuman NPC pipeline in Unreal Engine 5.7, powered entirely on-device using NVIDIA ACE.
The demo’s Action and Cut workflow enables NPCs to perform dynamically in-character during gameplay, then break character to collaborate directly with developers by refining motivations, generating backstory material, and helping build future training data.
To stay up to date with the latest NVIDIA game development news, check out the full list of game developer resources:
- Subscribe to our newsletter (select gaming as your industry)
- Follow us on social: X, LinkedIn, Facebook, and YouTube
- Join our Discord community
Tags
About the Authors
Phillip Singh is a senior developer marketing manager at NVIDIA, specializing in games and consumer AI. He is dedicated to empowering the next generation of game developers through neural rendering and AI graphics technology. Phillip earned his master's in business administration from the Santa Clara University Leavey School of Business.
Comments
More from NVIDIA Developer Blog
-
Develop High-Performance GPU Kernels in C++ with NVIDIA CUDA Tile
May 26
-
NVIDIA CUDA 13.3 Enhances GPU Development with Tile Programming in C++, Compiler Autotuning, and Python Updates
May 26
-
Extract More Kernel Performance with NVIDIA CompileIQ Auto-Tuning
May 26
-
Run Key Genomics and Protein Folding Workloads Faster with NVIDIA RTX PRO 4500 Blackwell
May 26
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.