Pistachio runs as a headless daemon or CLI on any server. You stake MOR, manage sessions automatically, and expose an OpenAI-compatible API to your users. They send prompts, you route them to Morpheus providers. You're the concierge — handling blockchain complexity so your users don't have to.
Your users connect to your gateway endpoint. You connect to the Morpheus network via Pistachio. The AI responses flow through peer-to-peer — no centralized API, no data logging, no single point of failure. You're a value-add service vendor on top of decentralized infrastructure.
Any Linux/Mac server, VPS, or cloud instance. Runs headless — no GUI needed.curl -fsSL https://raw.githubusercontent.com/drm3labs/drm3-releases/main/products/pistachio/install.sh | sh
MOR for staking sessions. ETH for gas on restakes. Your MOR is returned after each session — only gas is consumed.
Every wallet gets 20 MOR baseline, plus a free +20 MOR on-chain claim (1–2 models). Operators need Gold or Diamond for multi-concurrent lanes across many models.
Pistachio auto-manages sessions — staking, restaking, provider selection, failover. You issue API keys to your users.
They see an OpenAI-compatible API. They send prompts. Responses come from Morpheus providers. They never touch MOR, ETH, or a wallet.
Add AI to your product without OpenAI dependencies. Point your backend at your Pistachio instance. No vendor lock-in.
Offer private, censorship-resistant AI to your organization. Data never leaves the peer-to-peer connection.
Run a multi-model AI platform. Issue API keys. Charge per request or per subscription. You're the middleman that adds value.
Abstract away blockchain for non-technical users. They pay you in fiat, you handle MOR staking and session management behind the scenes.