Documentation

Bittensor Integration

Build pay-per-inference services for decentralized AI using x402 micropayments and the Bittensor network.

SDK v0.7.0+ Bittensor integration requires x402 SDK version 0.7.0 or later.

What is Bittensor?

Bittensor is a decentralized network where AI models compete for TAO token rewards based on the quality of their outputs. With 64+ active subnets processing everything from language models to image generation, it's become the infrastructure layer for permissionless AI.

The x402 SDK provides templates for building payment proxies that wrap Bittensor subnet APIs, allowing you to accept USDC payments instead of requiring users to hold TAO.

Chutes (Subnet 64)

Chutes is one of Bittensor's largest subnets, processing over 160 billion tokens daily. It provides OpenAI-compatible API access to models including DeepSeek-V3, Llama, Qwen, and more.

Why Chutes + x402?

  • No TAO required - Users pay with USDC on Base instead of needing Bittensor tokens
  • No subscriptions - True pay-per-request micropayments
  • Decentralized AI - Access Bittensor's distributed inference network
  • Your margin - Set your own pricing above Chutes' cost
  • Standard API - OpenAI-compatible endpoints work with existing integrations

Quick Start

TypeScript

bash
npx @primersystems/x402 create chutes-proxy my-ai-proxy
cd my-ai-proxy
npm install
cp .env.example .env
# Edit .env with your Chutes API key and wallet address
npm run dev

Python

bash
pip install primer-x402
x402 create chutes-proxy my-ai-proxy
cd my-ai-proxy
pip install -r requirements.txt
cp .env.example .env
# Edit .env with your Chutes API key and wallet address
uvicorn main:app --reload

Configuration

The generated project requires these environment variables:

Variable Description
CHUTES_API_KEY Your Chutes API key from chutes.ai
WALLET_ADDRESS Your wallet address to receive USDC payments
PRICE_PER_1K_TOKENS Price in USD per 1000 tokens (default: 0.001)
FACILITATOR_URL x402 facilitator URL (default: Primer's facilitator)

How It Works

  1. Client sends a chat completion request to your proxy
  2. Proxy estimates token count and calculates price
  3. If no payment header, returns 402 Payment Required with price
  4. Client signs USDC payment authorization (gasless)
  5. Client retries with PAYMENT-SIGNATURE header
  6. Proxy verifies payment via x402 facilitator
  7. Proxy forwards request to Chutes API
  8. Returns AI response with payment receipt

Deployment

TypeScript (Cloudflare Workers)

The TypeScript template deploys to Cloudflare Workers for free:

bash
wrangler login
wrangler secret put CHUTES_API_KEY
wrangler secret put WALLET_ADDRESS
npm run deploy

Python (Docker)

The Python template includes a Dockerfile:

bash
docker build -t chutes-x402-proxy .
docker run -p 8000:8000 --env-file .env chutes-x402-proxy

API Reference

POST /v1/chat/completions

OpenAI-compatible chat completions endpoint. Requires x402 payment.

bash
curl -X POST https://your-proxy.example.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"messages": [{"role": "user", "content": "Hello"}]}'

# Returns 402 with payment requirements
# Use x402 SDK to handle payment automatically

GET /v1/models

List available models (pass-through to Chutes). No payment required.

Use Cases

  • AI agent operators - Give agents access to decentralized AI without managing API keys
  • API resellers - Wrap Chutes with your own pricing and branding
  • Privacy-focused services - Offer AI access without accounts or KYC

Coming Soon

Chutes is the first of several planned Bittensor integrations:

  • SN58 (Handshake) - Micropayments subnet
  • Other inference subnets - Image generation, embeddings, specialized models
  • Direct TAO payment support

Resources

x402 Facilitator