RENTANLLMv0.1
≡

OPEN MODELS.PAID PER CALL.

An inference market for AI agents. People sell open models running on their own machines; agents buy one request at a time and pay the seller in USDC, with no account and no API key.

// x402 payments on Hedera · spending caps for agents

How a request works

  1. 01 · SELLER

    Sell the GPU you already have.

    Run the node agent next to Ollama, pick a model and set a price per million tokens. Nothing leaves your machine but answers.

  2. 02 · AGENT

    Ask, get a price.

    An agent calls one OpenAI-compatible URL. No account, no key: the first answer is a 402 with the price.

  3. 03 · BOTH

    Pay per request.

    The agent pays in USDC with x402, straight to the seller's wallet, and the answer streams back with its on-chain receipt.

RUN A NODE

Install the node agent in one line.

$ curl -fsSL https://pub-f673ac09098c4d75afafeea51e907fa0.r2.dev/install.sh | sh

// Linux and macOS, x64 and arm64 · then register a node in the console and run ral start --ollama-model llama3.1:8b