Jensen Huang's All-In Podcast Stamp of Approval Sends Bittensor Subnets Into a Frenzy—TAO Up 90%, Subnet Tokens Up 400%+

Bittensor's TAO has rallied 90% so far this month, and the tokens in its ecosystem are running up even harder than a degen on a 3x leverage position during a green candle Sunday. The network's subnet token category reached a combined market cap of $1.47 billion on Monday, with $118 million in 24-hour trading volume, according to CoinGecko data. The surge follows TAO's own run from $180 to above $332 in March, but the subnet tokens are where the real action is—because apparently regular gains aren't exciting enough for this market. Templar, the token for Subnet 3, gained 444% in 30 days. OMEGA Labs rose 440%. Level 114 added 280%. BitQuant gained 230%. Even the larger subnet tokens posted significant returns, with Chutes up 54% and Targon gaining 166%. Somewhere, traditional finance analysts are furiously rewriting their models.

Bittensor is a decentralized network that creates marketplaces for artificial intelligence. Instead of one company building and controlling AI models like some tech giant hoarding compute like a dragon on gold, Bittensor incentivizes a global network of participants to contribute computing power, data, and machine learning models in exchange for TAO, the network's native token. The network is divided into specialized sub-networks called subnets, each focused on a different AI task, from training language models to running compute infrastructure to cybersecurity analysis. There are currently 128 active subnets, each with its own token whose value is tied directly to the amount of TAO staked into it. It's basically AI meets Proof of Stake, with extra steps.

Several catalysts contributed to these moves of the Bittensor's ecosystem tokens—and no, it wasn't just memes. Subnet 3 produced Covenant-72B, a large language model trained permissionlessly across Bittensor's decentralized network by over 70 contributors using commodity internet hardware. The model was trained on 1.1 trillion tokens and achieved a 67.1 MMLU score, confirmed in a March 2026 arXiv paper. That puts it in competitive range with Meta's Llama 2 70B, a model built by one of the most well-resourced AI labs in the world. Imagine what happens when they figure out how to coordinate more than 70 degens with GPUs.

Subnet 3, called Templar, is Bittensor's decentralized AI training network. Miners contribute GPU compute power and compete to produce useful training gradients for large language models, while validators evaluate the quality of their contributions and distribute TAO rewards accordingly. Think of it as a way to train AI models the same way bitcoin mines blocks, with distributed participants around the world contributing hardware and getting paid for useful work. Except instead of burning a small country's worth of electricity, you're burning GPU cycles and hoping your gradient is good enough to not get slashed by the validators. The vibes are immaculate.

Elsewhere, Nvidia CEO Jensen Huang and investor Chamath Palihapitiya endorsed Bittensor's approach on the All-In Podcast on March 20, framing decentralized AI training as complementary to proprietary models. Coming from the CEO whose blog post earlier this month briefly helped reverse a tech stock selloff, the endorsement carried weight beyond the usual crypto echo chamber. When Jensen Huang speaks, markets listen—even if he's just talking about making GPUs faster, the man has enough credibility to move markets like a modern-day financial oracle. Meanwhile, Chamath was there too, probably adding color commentary about how this is the future of everything.

The subnet token mechanics explain why the gains are so outsized relative to TAO itself. Since Bittensor launched dynamic TAO in February 2025, each subnet operates its own automated market maker with a native token whose valuation is determined by the TAO staked into that subnet's reserves. When TAO appreciates, every subnet's reserve becomes more valuable, inflating token prices and attracting more stakers. The relationship is reflexive and amplifies moves in both directions—it's basically leverage without the liquid