BanklessTimes
Home Articles Nvidia to Launch Open Source AI Agent Platform

Nvidia to Launch Open Source AI Agent Platform

Simon Simba
Simon Simba
Simon is a writer with five years experience in crypto and iGaming. He currently works as a freelance writer at BanklessTimes where he focuses on simplifying daily crypto developments for readers. He discovered crypto in 2022 while writing news about NFTs for a news website in the US, and has since written for two other international NFT projects, and a Web3 gaming agency.
Updated: March 10th, 2026
Editor:
Joseph Alalade
Joseph Alalade
Editor:
Joseph Alalade
News Lead and Editor
Joseph is a content writer and editor who has actively participated in crypto for over 6 years. He enjoys educating others about Web3 and covering its updates, regulatory developments, and exciting stories.

NVIDIA is preparing to launch an open-source platform that helps companies build and run teams of AI agents. The project will give developers tools to connect different AI systems, watch their behavior, and improve speed and cost.

What the Nvidia AI Agent Platform Does

The planned platform builds on Nvidia’s existing NeMo Agent Toolkit and Agent Intelligence (AIQ) libraries. These open-source tools already let developers link many agents and tools, even if they use different frameworks.

NVIDIA’s toolkit treats agents and tools like simple functions that can be reused in many workflows. This design helps teams combine older systems with new AI models without rebuilding everything from scratch.

Reports say that Nvidia is trying to sell the new platform, which is currently called “NemoClaw,” to companies that make business software. Companies might use AI agents within their current systems to help workers with tasks such as searching, reporting, or customer service.

NVIDIA envisions the platform remaining open source and not requiring NVIDIA GPUs, despite its optimization for them. It also wants to add security and privacy capabilities, which are critical for companies handling sensitive data.

Link to Nvidia’s Open Models

The agent platform aligns with Nvidia’s broader push into “agentic AI” through its Nemotron 3 family of open models. These models are designed for reasoning, tool use, and multi-step workflows, which fit AI agents that must plan and act.

NVIDIA has also made open datasets and training tools available, such as NeMo Gym and NeMo RL, to help teams build safer, more reliable agents. The open models, data, and agent toolkit form a stack that developers can review, test, and modify.

NVIDIA will probably give additional information about the platform at its next developer event in San Jose. The startup is already in talks with major tech companies about collaborating on the agent platform.

More businesses may move from simple chatbots to coordinated AI agents that perform real tasks if Nvidia builds a strong open-source agent platform. This shift could make AI appear more like a small team of digital specialists rather than a single assistant.

READ MORE: Hyperliquid’s HYPE Price Jumps as Top Expert Sees a 330% Surge

Follow Bankless Times on Google News

We`ve got crypto covered – every trend, every insight, every move that matters. Add us to your feed and stay ahead of the market.

Contributors

Simon Simba
Simon is a writer with five years experience in crypto and iGaming. He currently works as a freelance writer at BanklessTimes where he focuses on simplifying daily crypto developments for readers. He discovered crypto in 2022 while writing news about NFTs for a news website in the US, and has since written for two other international NFT projects, and a Web3 gaming agency.