NVIDIA launches G-Assist plug-in hackathon to empower RTX system customization

NVIDIA invites developers to build custom plug-ins for its experimental G-Assist artificial intelligence assistant, with prizes for top projects in its July hackathon.

NVIDIA is inviting developers and enthusiasts to participate in its ´Plug and Play: Project G-Assist Plug-In Hackathon´, a virtual competition running through Wednesday, July 16. The event centers on Project G-Assist, NVIDIA´s experimental artificial intelligence assistant for GeForce RTX systems, which is accessible through the NVIDIA App. G-Assist leverages a compact language model running on-device to enable real-time system tuning, GPU control, and optimization using conversational commands, all without leaving gameplay or switching programs.

Hackathon participants are challenged to extend G-Assist´s capabilities by building custom plug-ins. These plug-ins can be created in Python for rapid prototyping, in C++ for performance-focused applications, or through system-specific automations for deeper integration. NVIDIA also offers a no-code or low-code ChatGPT-based Plug-In Builder, making plugin creation accessible to broader audiences. To enter, developers must submit their source code via a GitHub repository, prepare a demonstration video, and promote their plugin using the #AIonRTXHackathon hashtag across major social media channels. Submissions are evaluated on innovation, technical execution, and community usability, with winners announced on August 20.

The first prize is a GeForce RTX 5090 laptop, with RTX 5080 and 5070 GPUs awarded for second and third place, respectively. Winners also will be featured on NVIDIA channels, connect with the G-Assist team, and receive NVIDIA Deep Learning Institute course credit. NVIDIA provides ample support and resources, including sample plugins—such as integrations for Google Gemini, Discord, Spotify, Twitch, and IFTTT—detailed documentation, and a lively Discord channel for community help. A dedicated webinar on July 9 covers plugin fundamentals, with further insights available through NVIDIA´s GitHub and technical blogs. To participate in Project G-Assist, a GeForce RTX 50, 40, or 30 Series GPU with at least 12GB VRAM, supported operating system, compatible CPU, and recent drivers are required.

54

Impact Score

Scammers use Artificial Intelligence to pose as UK family businesses

Foreign firms are using Artificial Intelligence-generated images and fabricated back stories to impersonate family-run UK boutiques and lure shoppers, often via social media ads. Customers report poor-quality goods, high return costs and misleading claims about where companies are based.

What Meta’s purchase of Rivos says about RISC-V

Meta has acquired Rivos, a startup that taped out a CUDA-compatible RISC-V processor, signaling that Nvidia’s grip on Artificial Intelligence infrastructure may face new pressure from open architectures.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.