Bittensor's TAO token surged past $300 for the first time since November on March 20, 2026, following an unexpected endorsement from Nvidia CEO Jensen Huang. During an All-In Podcast episode with venture capitalist Chamath Palihapitiya, Huang compared Bittensor to Folding@Home, the early 2000s volunteer computing project, praising its decentralized approach to AI training. The comments triggered a 28% rally, pushing TAO from $243.50 to $310.60 before settling around $298-$304.
Trading volume exploded to $677 million daily, with open interest reaching $361 million at multi-month highs, signaling heightened institutional and speculative interest in decentralized AI infrastructure.
The Podcast Moment That Moved Markets
The catalyst emerged when Palihapitiya highlighted Bittensor's Subnet 3 achievement: training a 4-billion-parameter Llama model using distributed compute from random participants worldwide. Each contributor provided excess processing power and earned a share of rewards while maintaining coordinated progress—a 'pretty crazy technical accomplishment,' as Palihapitiya described it.
Huang responded by calling Bittensor 'a modern version of folding@home,' referencing the landmark distributed computing project that harnessed idle PCs for protein folding research. He argued the future of AI will blend proprietary systems with open, decentralized networks rather than favoring one approach exclusively. While Huang didn't explicitly promote the token, his technical validation from AI's hardware leader sparked immediate market reaction.
What Makes Bittensor Different
Launched in 2021, Bittensor operates as a decentralized machine learning network where participants compete in specialized subnets—AI marketplaces where miners earn TAO tokens by delivering superior models or inferences. This structure creates an open-source alternative to centralized AI labs, addressing concerns about data monopolies and computational centralization.
The network recently deployed Covenant-72B on Subnet 3, marking the largest decentralized pre-training initiative documented to date with 72 billion parameters. This milestone, combined with the Dynamic TAO (dTAO) upgrade, shifted TAO from a pure governance token to one with utility in staking and collateral, enhancing miner liquidity. Subnet-specific tokens like τemplar (Templar) surged 194% weekly as the ecosystem expanded.
Bittensor is now scaling from 128 to 256 subnets, broadening its AI marketplace capabilities and deepening its technical infrastructure.
Institutional Validation Accelerates Momentum
Beyond the podcast endorsement, institutional developments amplified TAO's rally. Grayscale's TAO Trust received SEC-reporting status in March 2026, providing regulated exposure similar to a spot ETF. South Korean exchange Upbit also listed TAO, improving price discovery and liquidity for the token.
The combination drove TAO to 46% monthly gains and 7.2% year-to-date returns, even as broader cryptocurrency markets faced headwinds from Federal Reserve rate stability. Open interest climbed to $361 million while subnet tokens rallied alongside the main token, demonstrating ecosystem-wide strength.
Decentralized AI Gains Credibility
Huang's comments signal growing mainstream recognition for decentralized AI infrastructure as a viable complement to centralized systems. His vision of coexistence between proprietary and open models validates Bittensor's approach, potentially attracting capital from investors following Nvidia's AI narrative.
The network now faces the challenge of sustaining technical momentum against competitors while navigating cryptocurrency market volatility. Analysts identify $285-$310 as key resistance levels, with subnet expansion and continued model training milestones likely to drive the next phase of growth.
For investors and developers alike, Bittensor's rally demonstrates that decentralized AI networks have moved from theoretical promise to production-grade infrastructure—with mainstream tech leaders taking notice.