At Gamescom, NVIDIA announced its first significant update to Project G-Assist, the experimental on-device AI assistant that lets gamers and creators tune their RTX systems with simple voice or text commands.
The update introduces a more innovative AI model that uses 40% less VRAM, improves tool-calling intelligence, and now supports all RTX GPUs with 6GB or more VRAM, including laptops. A new G-Assist Plug-In Hub also makes it easy for users to discover, download, and install plug-ins that unlock even more functionality.
NVIDIA also revealed an upcoming path-traced particle system for the RTX Remix modding platform, arriving in September. This feature brings fully simulated physics, realistic reflections, and dynamic shadows to classic games.
Finally, the company named the winners of the NVIDIA and ModDB RTX Remix Mod Contest, spotlighting the most creative uses of RTX Remix.
G-Assist Gets Smarter and More Accessible
Modern PCs are incredibly powerful, but managing settings across software, GPUs, and peripherals can feel overwhelming. Project G-Assist simplifies this, acting as a central AI-powered command hub. With voice or text, users can.
Run diagnostics to optimize game performance.
Monitor frame rates, latency, and GPU temperatures
Adjust GPU and peripheral settings (like keyboard lighting)
The new lightweight model ensures faster responses while consuming less memory, extending G-Assist to more RTX systems, including laptops. Getting started is easy,
Install the NVIDIA app and the latest Game Ready Driver (Aug. 19 release).
Download the G-Assist update.
Press Alt+G to activate.
In September, another update will add laptop-specific commands like NVIDIA BatteryBoost and Battery OPS.
Introducing the G-Assist Plug-In Hub
In partnership with mod.io, NVIDIA is launching the G-Assist Plug-In Hub, a one-stop shop for discovering, downloading, and managing plug-ins. Users can simply ask G-Assist which plug-ins are available and install them via natural language commands.
Highlights from the G-Assist Plug-In Hackathon include.
Omniplay: lets gamers access lore or take in-game notes from online wikis.
Launchpad: launch and toggle custom app groups on the fly.
Flux NIM Microservice: generate AI images directly within G-Assist using NVIDIA NIM.
Winners of the hackathon will be announced on Aug. 20. Building custom plug-ins is also straightforward, utilizing JSON, Python scripts, and the G-Assist Plug-In Builder, which supports natural language coding.
RTX Remix: Breathing New Life Into Classics
NVIDIA RTX Remix empowers modders to reimagine classic PC titles with cutting-edge graphics. Since launch, the community has created 350+ projects and released over 100 mods, totaling more than 2 million downloads.
At Gamescom, NVIDIA announced the winners of the RTX Remix Mod Contest.
Best Overall & Best Use of RTX: Painkiller RTX Remix (by Binq_Adams)
Runner-Up: Vampire: The Masquerade – Bloodlines RTX Remaster (by Safemilk)
Most Complete Mod: Painkiller RTX Remix (by Binq_Adams)
Runner-Up: I-Ninja Remixed (by g.i.george333)
Community Choice: Call of Duty 2 RTX Remix of Carentan (by tadpole3159)
Modders combined RTX Remix with generative AI tools like ComfyUI to enhance textures, generate assets, and rebuild environments. The results, ranging from Gothic horror atmospheres to polished modern visuals, highlight the power of RTX technologies.
Another RTX Remix update in September will add support for creating fully path-traced particle effects, unlocking new possibilities for over 165 compatible games.
Get Started
Download G-Assist and RTX Remix directly from the NVIDIA app.
Explore contest-winning mods on ModDB or join the RTX Remix Showcase Discord for sneak peeks at in-progress projects.
Follow the RTX AI Garage blog series to learn more about NVIDIA NIM microservices, AI workflows, and community-driven innovations.
Stay connected with NVIDIA on Facebook, Instagram, TikTok, X, and Discord, and subscribe to the RTX AI PC newsletter for the latest updates.
Join the conversation! Your thoughts help the community grow.