Nvidia Is Reportedly Growing Its Personal Reply to OpenClaw

Nvidia is on the verge of asserting its personal Claw, according to Wired. Are you able to consider it? An Nvidia claw!

Which claw do you employ in your agentic AI duties? The light-weight Nanoclaw? The safety-focused IronClaw? Oh, I can inform out of your trend sense you’re retro, and you favor the O.G.—OpenClaw. Typically I get nostalgic for six weeks in the past too. Easier occasions!

For those who don’t know what the hell I’m speaking about, you’re not a wicked AI freak, which is lucky for you. Please remember, nevertheless, that this complete “claw” pattern is transferring in a short time. Nvidia’s place because the premier developer of AI chip architectures, and of CUDA, the underlying proprietary software program platform behind a lot of the AI world, might imply Nvidia is seeking to set requirements for an essential new tech class by stepping into the claw recreation.

Purchased a brand new Mac mini to correctly tinker with claws over the weekend. The apple retailer particular person advised me they’re promoting like hotcakes and everyone seems to be confused 🙂

I am positively a bit sus’d to run OpenClaw particularly – giving my non-public knowledge/keys to 400K strains of vibe coded…

— Andrej Karpathy (@karpathy) February 20, 2026

Claws, a {hardware} and software program pattern that began with the discharge of OpenClaw (previously often known as Clawdbot and Moltbot) final 12 months, are usually wrappers for LLMs like Claude Code and OpenAI’s Codex that ostensibly operate as private assistants that may carry out duties that contain writing code and shopping the web. Customers usually arrange a devoted pc to run a claw, plug an costly LLM subscription into it, give it entry to their private knowledge and accounts, after which talk with it over a messaging app like WhatsApp (Claws are additionally, famously, a security nightmare).

The creator of OpenClaw, Austrian software engineer and former entrepreneur Peter Steinberger, was employed by OpenAI final month. OpenAI CEO Sam Altman wrote that his mission at OpenAI is “to drive the subsequent era of private brokers,” and that he expects that what Steinberger creates there’ll “rapidly develop into core to our product choices.”  

In response to Wired, whose reporting on Nvidia comes from nameless leaks—or “individuals acquainted with the corporate’s plans,” to make use of Wired’s phrase—Nvidia has been approaching enterprise software program firms to debate it’s claw platform, which is known as (for now?) NemoClaw. Enterprise software program platforms have been subjected to an all-out inventory value assault from traders currently, whose market behavior suggests they believe the software-as-a-service (SaaS) model is overvalued attributable to impending automation made potential by instruments like OpenClaw.

Nvidia is seemingly permitting these firms free early entry to NemoClaw, that means they will supply precise work duties from claw-style AI brokers, whether or not their techniques run on Nvidia chips or not, in trade for contributions to Nvidia’s claw challenge. Google, Adobe, Salesforce, Cisco, and CrowdStrike, are the potential companion firms talked about by Wired, though they’ve been silent to date on whether or not or not they’re partnering with Nvidia. 

NemoClaw is reportedly open-source, and its moniker means that it’s meant to be powered by the Nemotron household of open-source fashions, like Nemotron 3 which was introduced final 12 months. The press release for Nemotrom 3 says these fashions are “designed to energy clear, environment friendly and specialised agentic AI improvement throughout industries.” 

In response to Wired’s reporting, NemoClaw will probably be introduced at Nvidia’s GTC developer convention subsequent week. That may imply, if the Wall Street Journal’s reporting is correct, that it’s going to coincide with the discharge of a new inference chip Nvidia plans to release as well.