Bittensor is building a confidential routing layer that could reshape AI inference pricing, delivery, and verification. The integration focuses on Subnet 28, internally called "gm," which uses Trusted Execution Environments to let miners compete on routing costs, payments, and model access while keeping user queries private.
Subnet 28 creates a privacy-first marketplace for AI inference. Trusted Execution Environments (TEEs) are hardware-level secure enclaves that process data without exposing it to the machine's operator. Even the miners running AI queries cannot see what users are asking. The system generates verifiable attestations-cryptographic proof that the request was handled inside a secure enclave without tampering. Miners compete not just on speed or quality but on routing costs and payment terms, creating a dynamic pricing layer on top of Bittensor's existing subnet economy.
Development discussions took place between May 28 and 29, 2026. The architecture has been compared to "OpenRouter or Venice AI, but on Bittensor," positioning centralized inference aggregators as the competitive benchmark.
Bittensor's Chutes subnet (SN64) processes between 100 billion and 120 billion tokens per day. Between 20% and 25% of that traffic routes through OpenRouter, making Chutes one of the platform's top inference providers. Other subnets including Targon and Nineteen are also operational on OpenRouter. OpenRouter itself raised $113 million in May 2026, led by CapitalG, Google's independent growth fund, at a $1.3 billion valuation.
Users accessing Bittensor-powered models through OpenRouter do not need TAO tokens; they interact through OpenAI-compatible APIs.
For investors, OpenRouter's $1.3 billion valuation signals strong enterprise demand for multi-model routing. Bittensor subnets competing as top providers demonstrate decentralized inference is commercially viable. The TAO token can capture value at the infrastructure layer through staking and validator incentives, even as end users avoid holding the token.
Key challenge: TEE-based confidential routing must maintain throughput levels while adding encryption and attestation overhead. Whether miners can absorb these costs while remaining price-competitive will determine adoption speed.