Bittensor (TAO) is a decentralized AI network where participants supply AI models and compute and earn TAO based on their contributions.
Bittensor replaces single-company control of AI development with an open system in which participants worldwide contribute models, data and computation, and the network scores that work and distributes TAO accordingly. The project has been led by the Opentensor Foundation, founded by Jacob Steeves and Ala Shaabana.
Bittensor presents itself as an alternative to AI built by large technology companies, aiming for AI infrastructure that anyone can join and own. Bittensor is often cited as one of the leading projects combining crypto and AI.
Bittensor is split into many specialized markets called subnets, and each subnet focuses on a specific task such as text generation, images, data collection or prediction. Inside a subnet, miners compete to deliver better outputs.
Bittensor validators score the quality of miner outputs, and higher-scoring contributions receive more TAO rewards. This market of competition and evaluation designed to raise AI quality is the core mechanism of Bittensor.
Data provided by CoinGecko
TAO follows a supply design similar to Bitcoin, with a maximum supply capped at 21 million tokens according to the project documentation. New TAO enters circulation only as rewards for network contributions, with no pre-sale.
Bittensor's first halving took place in December 2025, cutting daily issuance from about 7,200 TAO to about 3,600 TAO according to multiple news reports. As of September 2026, industry estimates place the next halving around 2029.
The TAO scarcity design means rewards for AI contributions are paid in a coin with a limited supply. A capped supply by itself, however, does not guarantee any value.
Bittensor activated the dTAO (Dynamic TAO) upgrade in February 2025, giving each subnet its own alpha token. Since dTAO, participants stake TAO into specific subnets, and market demand determines how rewards are allocated across subnets.
Under dTAO, capital and rewards flow toward subnets that perform well and attract demand, while subnets with little demand see their rewards shrink. The Bittensor ecosystem has expanded on this basis into a bundle of markets spanning text, images, prediction, data and other AI fields.
The Bittensor roadmap focuses on increasing the number and variety of subnets and refining the contribution, evaluation and reward mechanisms. Ongoing updates also aim to tie subnet rewards more closely to real demand and measurable performance.
Bittensor treats getting outside apps and businesses to actually use subnet-produced AI services as a key task. A central question is whether decentralized AI on Bittensor can move beyond research into services that generate real revenue.
The biggest challenge for Bittensor is that the value of AI contributions is hard to evaluate objectively. Judging whether validator scoring is fair and whether rewards reach genuinely useful AI is difficult, and some observers note that the subnet and dTAO structure is complex for beginners.
The TAO price has swung widely with sentiment around the AI theme, and a high share of staked tokens leaving a thin tradable float is also cited as a source of volatility. This article is for information only and is not investment advice; always make investment decisions at your own responsibility.
Links to major exchanges below. Check listing and fees on each exchange. Not investment advice; links may contain affiliate (referral) codes.