gm Subnet Launches with TEE-Enabled Model Routing

gm (SN28) went live on May 28th as an OpenRouter/Venice-like marketplace for LLM inference on Bittensor, featuring Trusted Execution Environment (TEE) capabilities to ensure data privacy and model integrity. The subnet leverages existing API providers (Claude, GPT, Gemini) and open-source models via TEE partners like Chutes, Near, and Phala, rather than building its own GPU infrastructure. Initial payment support includes USDC and bank cards; x402 payment integration planned post-launch.

• •TEE ensures gateway and miners cannot view data in transit
• •TEE validates miners run correct model, preventing swaps
• •USDC and bank cards enabled at launch; x402 roadmapped
• •Model routing leverages existing providers, not proprietary GPU stack