Bittensor Tokens Hit $1.5B as Jensen Huang Backs TAO
What to Know
- $1.47 billion — combined market cap reached by Bittensor subnet tokens on Monday, per CoinGecko data
- TAO rallied 90% in March 2026, climbing from $180 to above $332, but subnet tokens outpaced it dramatically
- Subnet token Templar gained 444% in 30 days after Bittensor's Subnet 3 released the Covenant-72B AI model
- Jensen Huang and Chamath Palihapitiya endorsed decentralized AI training on the All-In Podcast on March 20
Bittensor TAO has had a strong March — up 90% and counting — but the real fireworks are happening one layer deeper. The network's subnet token category crossed a combined market cap of $1.47 billion on Monday, with $118 million in 24-hour trading volume, according to CoinGecko. TAO's run from $180 to above $332 is the headline number, but the subnet tokens are trading like leveraged calls on the parent protocol — and some of them are up over 400% in a single month.
Why Subnet Tokens Are Outrunning TAO
The mechanics here aren't complicated once you understand how Bittensor subnet tokens actually work. Since the launch of dynamic TAO in February 2025, each of the network's 128 active subnets operates its own automated market maker. Every subnet token's valuation is pegged directly to the amount of TAO staked into that subnet's reserve pool. When TAO goes up, every reserve gets more valuable. That inflates token prices, which attracts more stakers, which pushes prices higher still. The relationship is reflexive — it amplifies gains and, when the tide turns, it amplifies losses just as fast.
Think of it like this: TAO is the base asset and subnet tokens are effectively leveraged bets on whether specific AI tasks running on the network are worth anything. The numbers from this past month back that up. 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, more established subnet tokens posted numbers most crypto traders would be thrilled with: Chutes up 54%, Targon up 166%. This is what happens when a reflexive token structure meets genuine fundamental news.
- Templar (Subnet 3): +444% in 30 days
- OMEGA Labs: +440%
- Level 114: +280%
- BitQuant: +230%
- Targon: +166%
- Chutes: +54%
Covenant-72B: The Model That Started the Templar Surge
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. Validators then assess contribution quality and distribute TAO rewards accordingly. It's the Bitcoin mining model applied to AI: distributed hardware, permissionless participation, paid for useful work.
What Templar produced in March was legitimately surprising. Covenant-72B decentralized AI — a 72-billion parameter large language model — was trained across the Bittensor network by over 70 contributors using commodity internet hardware, not data center infrastructure. The model was trained on 1.1 trillion tokens and achieved a 67.1 MMLU score, confirmed in a March 2026 arXiv paper. For context, MMLU — Massive Multitask Language Understanding — benchmarks AI models across 57 academic subjects. A score of 67.1 puts Covenant-72B in competitive range with Meta's Llama 2 70B, a model built by one of the most well-resourced AI labs on the planet.
That's the part that deserves more attention. A permissionless network of anonymous GPU contributors, using hardware you could find in any gamer's rig, trained a model that can hold its own against Meta. Whether or not the Templar token sustains its gains, that is a genuine proof of concept.
What Does the Jensen Huang Endorsement Actually Mean?
On March 20, Nvidia CEO Jensen Huang and investor Chamath Palihapitiya appeared on the All-In Podcast and endorsed Bittensor TAO's broader approach — framing decentralized AI training as complementary to, rather than in competition with, proprietary models. The timing matters. Huang's credibility in AI hardware is unmatched, and his words move markets. Earlier in March, a blog post from him briefly reversed a tech stock selloff. An endorsement of decentralized AI infrastructure from the CEO of the company supplying most of the world's GPU compute is not nothing.
Call it a convergence of narratives: a real technical milestone from Covenant-72B, a structural amplifier in the subnet token mechanics, and a celebrity validator in Jensen Huang — all hitting in the same 30-day window. Crypto rarely needs more than one catalyst to run. This one had three.
Digital Currency Group subsidiary Yuma is already contributing to 14 different subnets, which suggests the institutional money isn't treating this as a meme cycle. When DCG — the same group behind Grayscale — is actively deploying across the ecosystem, it's infrastructure positioning, not speculation. That context is worth keeping in mind if you're trying to figure out whether this is a sustained rotation or a peak-euphoria moment to fade.
What Comes Next for Bittensor's Ecosystem?
A few things are worth watching. The network is planning to expand from 128 to 256 active subnets later in 2026, which means another wave of token launches — and more opportunities for the reflexive mechanics to work in both directions. A potential regulatory decision on converting the Grayscale TAO Trust into a spot ETF could open institutional access by late 2026, adding a fresh demand vector beyond current retail and crypto-native buyers.
The harder question is whether Covenant-72B was a repeatable outcome or a one-off milestone that happened to land during a favorable macro window. Subnet tokens with individual market caps ranging from $1 million to $137 million against a parent Bittensor TAO market cap of roughly $3 billion still have room to run if the network keeps shipping — but that reflexive relationship cuts both ways, and TAO doesn't need a big correction to send the smaller subnet tokens into a much sharper pullback.
For now, the ecosystem has momentum, a legitimate model to point to, and a credible public backer. Whether Bittensor keeps delivering AI breakthroughs at this pace — or whether the Huang endorsement was peak attention — will define how this ends up.
Frequently Asked Questions
What are Bittensor subnet tokens?
Bittensor subnet tokens are native tokens for each of the network's specialized sub-networks, called subnets. Each subnet focuses on a specific AI task. Token valuations are determined by the amount of TAO staked into each subnet's reserve, making them effectively leveraged instruments tied to the parent TAO token's price.
Why did Bittensor TAO surge 90% in March 2026?
TAO's March 2026 rally was driven by multiple catalysts: the release of Covenant-72B, a decentralized large language model trained permissionlessly on Subnet 3, and a public endorsement from Nvidia CEO Jensen Huang and investor Chamath Palihapitiya on the All-In Podcast on March 20, 2026.
What is Covenant-72B and how does it relate to Bittensor?
Covenant-72B is a 72-billion parameter AI model trained permissionlessly across Bittensor's Subnet 3 (Templar) by over 70 contributors using commodity hardware. It was trained on 1.1 trillion tokens, scored 67.1 on the MMLU benchmark, and its release was confirmed in a March 2026 arXiv paper — placing it in range of Meta's Llama 2 70B.
How do Bittensor subnet token mechanics work with dynamic TAO?
Since dynamic TAO launched in February 2025, each subnet runs its own automated market maker. Token prices are directly tied to TAO staked in that subnet's reserve. When TAO appreciates, every reserve becomes more valuable, pushing subnet token prices higher in a reflexive loop that amplifies moves in both directions.
This article is for informational purposes only and does not constitute investment advice. Every investment and trading decision involves risk. Readers should conduct their own research before making any financial decisions.
Topics
Bittensor TAOBittensor subnet tokensCovenant-72B decentralized AITAO rallydecentralized AI trainingdynamic TAOTemplar subnetMilan Torres
Senior Analyst
Milan covers Bitcoin markets, macro trends, and institutional crypto adoption with a focus on data-driven analysis.
Latest Crypto News
Top Stories
Stay ahead of the market.
Crypto news and analysis delivered every morning. Free.
More from TheCryptoWorld
About the Author
Contributor
TheCryptoWorld Staff is a contributor at TheCryptoWorld.
View all contributors
Follow thecryptoworld.io on Google News to receive the latest news about blockchain, crypto, and web3.
Follow us on Google News
