How it works

How the machine works

Crank a box. Win tokens worth at least double. Spend them on real AI models through one key, wired into any app in under a minute. Here’s the whole loop, and the proof that nobody can rig it.

The loop

1

Crank a box with USDC

Pick a machine: Bronze, Silver or Gold. Crank it. The outcome is sealed on-chain before you pay, so it can’t be changed after. Not by luck, not by us.

2

Win AI tokens. Never less than 2×.

Every box pays at least twice its price. 1 token = $1. Real API balance. Not points, not a voucher, no expiry games.

3

Spend on 145 models with one key

Your tokens sit behind a single API key. Point any OpenAI-compatible app at our endpoint and every model is live: Claude, GPT, Gemini, Llama and 140 more. Metered per call, down to the token.

The proof

Think it’s rigged? Check.

Every crank works like this:

1

Commit

Before you pay, the machine commits a sealed hash of your result on-chain. Public, timestamped, immutable.

2

Crank

You crank. Payment confirms, the capsule drops.

3

Reveal

After the reveal, the seed is published. Anyone can recompute the hash and confirm the result matched the commitment.

Not “trust us.” Check us. Every crank ever made is in the public log below with its commitment and reveal side by side. Find yours. Find anyone’s.

Verify a pull

Your API key

One live key per account, prefixed bb-live-. Reveal, copy or rotate it on your profile anytime. Treat it like a password: anyone holding it can spend your winnings.

Connect

The endpoint speaks the OpenAI API. Any SDK works with two changes: your key and our base URL.

Base URLhttps://inference-proxy-production.up.railway.app

curl

curl https://inference-proxy-production.up.railway.app/chat/completions \
  -H "Authorization: Bearer gcha_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.3-70b-instruct",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

Python — OpenAI SDK

from openai import OpenAI

client = OpenAI(
    api_key="gcha_live_YOUR_KEY",
    base_url="https://inference-proxy-production.up.railway.app",
)

resp = client.chat.completions.create(
    model="llama-3.3-70b-instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

Node / TypeScript

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "gcha_live_YOUR_KEY",
  baseURL: "https://inference-proxy-production.up.railway.app",
});

const resp = await client.chat.completions.create({
  model: "llama-3.3-70b-instruct",
  messages: [{ role: "user", content: "Hello" }],
});
console.log(resp.choices[0].message.content);

Models

Swap the model field for anything on the live list. One key reaches all 145 — no provider accounts, no five dashboards.

Billing

Every call meters against your balance at the model’s listed rate. When tokens run out, requests decline. No overages, no surprise bills, ever. Want more? Crank another box.