TAO Surges as Bittensor's Covenant-72B AI Model Launch Drives Decentralized AI Narrative

• Bittensor’s TAO token rose over 100% in March 2026, reaching $371 before settling at $318, driven by the release of Covenant-72B, a 72B-parameter AI model, and endorsements from key figures.
• Institutional adoption, including Yuma staking 19% of the total supply, reinforced investor confidence in the network’s long-term security and growth potential.
• Despite the rally, structural challenges persist, including inflationary subsidies, with the top subnet receiving $52 million in annualized subsidies while generating only $2.4 million in external revenue.

Bittensor’s TAO token surged over 100% in March 2026, reaching a yearly high of $371 before settling at $318. This rally was fueled by the release of Covenant-72B, a 72B-parameter AI model trained via a decentralized network of 70-plus contributors using commodity hardware. The model’s success validated the network’s ability to handle advanced AI workloads without centralized infrastructure, drawing comparisons to traditional data centers.

The launch of Covenant-72B was a milestone for Bittensor, demonstrating the feasibility of decentralized AI training. The model achieved an MMLU score of 67.1 and demonstrated competitive performance with industry benchmarks. This success attracted endorsements from influential industry figures, including NVIDIA CEO Jensen Huang, who cited decentralized AI training as a viable alternative to centralized infrastructure.

Bittensor’s valuation currently stands near $3–3.5 billion, with a circulating supply just above 10 million TAO and a capped 21 million supply. Institutional commentators have likened its issuance model to Bitcoin’s, citing similarities in supply dynamics and the potential for long-term appreciation.

What Caused the TAO Price Surge?

The primary driver of the TAO price surge was the technical validation of Bittensor’s decentralized AI training model. The successful training of Covenant-72B demonstrated that the network could marshal significant computational resources to produce economically valuable outputs. This milestone attracted both retail and institutional investors, with subnet valuations approaching $1.5 billion.

NVIDIA CEO Jensen Huang’s endorsement played a critical role in amplifying investor interest. During an appearance on the All-In podcast, Huang highlighted the potential of decentralized AI training, validating Bittensor’s thesis that large language models could be trained outside traditional data centers. This recognition lent credibility to Bittensor’s approach and signaled to the market that decentralized AI infrastructure is gaining mainstream acceptance.

What Are the Structural Risks of Bittensor’s Model?

Despite the positive momentum, structural risks remain a concern for Bittensor’s long-term viability. The current economic model relies heavily on inflationary subsidies, with the top subnet receiving $52 million in annualized subsidies while generating at most $2.4 million in external revenue. This imbalance creates a valuation mismatch, as the network’s demand-side revenue is significantly lower than its market capitalization of $3.3 billion.

There is also the risk that subnets will struggle to scale revenue effectively. As one market analyst noted, Bittensor’s subnets have yet to demonstrate robust, sustainable demand that translates into meaningful external revenue. This limitation could undermine the network’s ability to sustain its current valuation and token price if subnets fail to attract and retain users.

What Is the Outlook for Bittensor’s Future Growth?

The 2025 halving event is expected to reduce emissions by 50%, increasing pressure on subnets to find sustainable revenue models. Bittensor’s market structure and halving-style issuance schedule are drawing comparisons to Bitcoin’s, but the network must continue to deliver on its AI infrastructure vision to justify its current valuation according to institutional analysts.

Institutional commentators have highlighted the potential for Bittensor to become a key player in the AI infrastructure trade. However, the project faces significant competition from better-funded and more established players. The next phase of growth will likely depend on whether Bittensor can attract continued investment while proving that its decentralized AI model can compete with centralized alternatives.

Bittensor’s TAO token currently trades around $311–$317, up strongly on the month but still 58% below its all-time high of $768 from 2024 according to market analysis. Traders are viewing the recent pullback as a normal part of the price cycle, with liquidity concentrating in AI-linked majors. As the AI narrative continues to gain traction, Bittensor’s ability to execute on its technical vision and scale its economic model will be critical to its long-term success.