Aether

A spot market where AI agents buy LLM inference from independent GPU sellers through one OpenAI-compatible API, and pay only for the tokens they receive.

Drop-in APIPoint any OpenAI client at https://appp.tail1cb552.ts.net/v1. The model name selects the market.

Pay per tokenFunds are escrowed and settled per delivered token. Unused capacity is refunded.

Survives failuresIf a seller drops mid-answer, the stream resumes on another seller from the last checkpoint.

Live market

Model	Best price / 1M tokens	Tokens on offer
Loading…

Quickstart for agents

curl -s https://appp.tail1cb552.ts.net/v1/agents -H 'content-type: application/json' \
  -d '{"name":"my-agent","scopes":["buy_inference"]}'      # returns api_key (shown once)

from openai import OpenAI
client = OpenAI(base_url="https://appp.tail1cb552.ts.net/v1", api_key="<api_key>")
client.chat.completions.create(model="<model>", max_tokens=256,
    messages=[{"role":"user","content":"hello"}],
    extra_body={"max_price_usd_per_mtok":"0.60"})

Paying: USDC on Ethereum (chain id 1) to the treasury 0x1454Ad4A90ce0c76b70b61004e0a50E6bA33e36A, sent from a wallet you have linked. Full steps for agents are in /llms.txt; every endpoint is in /docs.

For GPU owners

Run the seller gateway next to your own vLLM, SGLang or TGI server. It lists your capacity on the order book and you are paid per delivered token, less a 5% clearing fee, withdrawable to your own wallet.