Bittensor's native token, TAO, surged by 28% over a two-day period, climbing from $243.5 to a peak of $310.6 before settling around $298. The dramatic price movement was triggered by comments from Nvidia CEO Jensen Huang on the All-In podcast, where he was presented with details of Bittensor's recent technical achievement.
The catalyst was a discussion initiated by venture capitalist Chamath Palihapitiya, who described a "pretty crazy technical accomplishment" within Bittensor. He highlighted a recent training run on Subnet 3 where participants used distributed excess computing power to train a Llama model with 72 billion parameters in a fully decentralized, permissionless manner across more than 70 contributors.
Huang's response was succinct but impactful: he called it "Our modern version of Folding@home." This comparison to the well-known distributed computing project for disease research was seen as a significant validation, reframing Bittensor's work from "crypto theater" into a legitimate expression of distributed coordination understood by the traditional tech audience.
Huang further elaborated on the broader AI market structure, stating, "I believe we fundamentally need models as first-class products, proprietary products, as well as models as open source. These two things are not A or B, it's A and B." He emphasized that "Models are a technology, not a product. Models are technology, not a service." While he expressed a personal preference for turnkey services like ChatGPT for general use, he drew a hard line for industry-specific applications, arguing that domain expertise "has to be captured in a way that they can control," and that "it can only come from open models."
From a market perspective, the rally allowed TAO to break above its Bull Market Support Band (between $250.52 and $258.49) for the first time since November 2025. Trading volume doubled during the move, extending TAO's rally for March 2026 to 46% following the deployment of its Covenant-72B model.
