Table of Contents
Mentat has launched Mentat Lend, a new money market built for Bittensor that lets TAO holders supply liquidity and lets Alpha holders borrow TAO without selling their subnet exposure.
The protocol's first market is live for Score, Bittensor Subnet 44. According to Mentat Lend's announcement, users can already deposit TAO to earn lending yield or post staked SN44 Alpha as collateral to borrow TAO.
The launch adds a much-needed DeFi primitive to Bittensor economy. Shizzy Unchained, commenting on the launch, described it as infrastructure Bittensor had been missing and said the subnet economy is "just getting started."
Let’s dive into how it all works.
How Mentat Lend Works
Mentat Lend is structured as a subnet-specific lending market.
TAO suppliers deposit into an isolated pool, while borrowers post staked subnet Alpha as collateral and borrow TAO from that pool. Mentat says the collateral remains staked during the loan, allowing borrowers to keep earning subnet emissions and validator rewards while accessing liquid TAO.
That design addresses one of the central liquidity problems in Bittensor. Under dynamic TAO, staking into a subnet gives users Alpha exposure and directs emissions, but operators and holders who need TAO for expenses, development, liquidity management, or other uses often face a tradeoff between keeping the Alpha position and selling it. A lending market creates a third path by allowing holders to borrow against the position instead.
The first implementation is focused on Score's SN44 Alpha. Score is a computer vision subnet that has become one of the more visible examples of Bittensor being used outside crypto-native workflows. We previously covered Score's partnership with PwC, where its Vision AI models will help enterprise clients leverage their existing camera networks to trigger automations across physical environments. The team is also diving into the TCG boom:
Bart Hillerich
A Credit Layer for Bittensor's Subnet Economy
The core argument behind Mentat Lend is that Bittensor already prices subnet ownership, but has not had a native way to price credit.
dTAO, on one hand, gives each subnet a live market between TAO and Alpha. Staking into a subnet is both an allocation decision and a market action. Users exchange TAO for Alpha exposure, and unstaking reverses that path through the subnet's pool. That mechanism helps price conviction in a subnet's upside.
On the other hand, borrow rates price something different. They show what TAO lenders require to accept the risk of a specific Alpha asset as collateral. In Mentat's framing, lending rates work as a signal for subnet downside risk while also generating yield.
The protocol's interest-rate model starts from Bittensor's root staking yield, which can be thought of as the network's floor rate for TAO capital because lenders can otherwise stake TAO to root rather than lend it. Each pool then adds a premium driven by utilization and subnet-specific risk. When demand to borrow rises or liquidity becomes scarce, rates are designed to move higher.
That structure is important for both sides of the market. Lenders are compensated for supplying TAO into a pool with utilization, collateral, liquidation, and smart-contract risks. Borrowers get open-ended TAO liquidity, but their cost of capital floats with pool conditions.
Why the Design Is Bittensor-Native
Mentat Lend is built around Bittensor's EVM layer rather than a generic external collateral wrapper. The protocol says it uses Bittensor precompiles for staking, subnet price reads, storage queries, and EVM-to-Substrate address mapping.
The most important design choice is collateral pricing. Mentat says the protocol does not rely on external price feeds. Instead, it prices Alpha collateral from Bittensor's own subnet AMM data and uses chain-recorded price observations to reduce the risk of single-transaction manipulation. It's done this way because subnet Alpha markets can be thinner and more idiosyncratic than major crypto assets.
A lending protocol that, for example, naively marks Alpha at spot price times quantity could overstate what a large collateral position is actually worth. Mentat says its oracle and valuation logic account for AMM depth, so collateral is measured closer to what it could be sold for through the pool.
The protocol also includes liquidation and risk controls tailored to subnet assets. Each pool uses subnet-specific liquidation loan-to-value settings, a liquidation incentive factor, a capacity gate tied to AMM depth, and a settlement process for deregistration events.
Risk Controls and Audits
Mentat says the protocol's contracts have been audited by Burra Security. The announcement states that two external audit rounds were completed, covering the core protocol and a Balancer adapter upgrade, with findings closed.
The design still introduces risks that differ from root staking. Lenders face pool-specific exposure, including smart-contract risk, collateral depreciation, liquidation shortfalls, high utilization that can delay withdrawals, and deregistration risk if a subnet exits the active set. Borrowers face liquidation risk, floating rates, possible pool-level trimming if aggregate debt exceeds the protocol's capacity bounds, and settlement risk if collateral is affected by subnet deregistration.
These risks are common in credit markets, but they matter more in Bittensor because subnet Alpha liquidity and market depth can vary widely. Mentat's isolated-pool structure is meant to prevent one subnet's risk from directly spilling into another pool, while letting each market develop its own borrowing cost.
Sparking Bittensor's Financial Layer
Mentat Lend arrives to help lead the activation of Bittensor's DeFi landscape, which needs to become viable and widely used for the network to become a leader in crypto.
On Solana, for example, 3/5 of the top protocols are directly focused on increasing ecosystem liquidity (Kamino lend/borrow, Sanctum & bnSOL liquid staking). Jupiter then has Jupiter Lend to get itself into the mix as well. Put together, we're looking at close to $5b in TVL overall.
With all the technical milestones we continue to see on Bittensor, the increased media coverage, etc., the time is ripe to increase the capital efficiency of the ecosystem.
Mentat Lend has arrived at just the right moment.
