# tx402.ai — Agent-Native LLM Inference Gateway
# x402 protocol · USDC micropayments · No API keys · No accounts

> tx402.ai is an x402 payment gateway for EU-sovereign LLM inference.
> AI agents pay per-request with USDC on Base. No API keys, no accounts, no KYC.
> Wallet = authentication. EU-based, GDPR-compliant, zero data retention.
> Powered by Tensorix EU infrastructure — https://tensorix.ai

## Quick Start for AI Agents

### Using @x402/fetch + @x402/evm (recommended)
```typescript
import { wrapFetchWithPayment, x402Client } from "@x402/fetch";
import { ExactEvmScheme, toClientEvmSigner } from "@x402/evm";
import { createWalletClient, http } from "viem";
import { base } from "viem/chains";
import { privateKeyToAccount } from "viem/accounts";

const account = privateKeyToAccount("0xYOUR_PRIVATE_KEY");
const wallet = createWalletClient({
  account,
  chain: base,
  transport: http(),
});

const client = new x402Client();
const signer = toClientEvmSigner(wallet);
signer.address = account.address;
client.register("eip155:8453", new ExactEvmScheme(signer));

const fetchWithPayment = wrapFetchWithPayment(fetch, client);

const response = await fetchWithPayment(
  "https://tx402.ai/v1/chat/completions",
  {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      model: "deepseek/deepseek-v3.2",
      messages: [{ role: "user", content: "Hello" }],
      max_tokens: 512,
    }),
  },
);
const data = await response.json();
```

### x402 Flow (automatic with wrapFetchWithPayment)
1. POST /v1/chat/completions → 402 response with USDC price + nonce
2. SDK signs USDC payment for quoted amount (PAYMENT-SIGNATURE header)
3. SDK retries with payment header → 200 + LLM response

> With wrapFetchWithPayment, steps 2-3 happen automatically.
> The agent just calls fetchWithPayment() once — the SDK handles the rest.

### API Endpoints
- GET  /                     Gateway info (JSON) or landing page (HTML)
- GET  /health               Health check
- GET  /v1/models            Available models + pricing + aliases (OpenAI-compatible)
- POST /v1/chat/completions  Chat completions (x402 gated)
- POST /v1/completions       Text completions (x402 gated)
- POST /v1/embeddings        Embeddings (x402 gated)
- GET  /openapi.json         OpenAPI 3.1 specification
- GET  /llms.txt             This file
- GET  /robots.txt           Crawler directives
- GET  /.well-known/ai-plugin.json  AI plugin manifest
- GET  /.well-known/x402     x402 protocol discovery (models, pricing, treasury)
- GET  /.well-known/openapi.json    OpenAPI spec (alternate path)

### Model Aliases (use these for convenience)
  minimax → minimax/minimax-m2.5
  m2 → minimax/minimax-m2
  m2.1 → minimax/minimax-m2.1
  m2.5 → minimax/minimax-m2.5
  deepseek → deepseek/deepseek-v3.2
  deepseek-chat → deepseek/deepseek-chat-v3.1
  deepseek-r1 → deepseek/deepseek-r1-0528
  glm → z-ai/glm-5
  glm-5 → z-ai/glm-5
  glm-4 → z-ai/glm-4.7
  kimi → moonshotai/kimi-k2.5
  moonshot → moonshotai/kimi-k2.5
  llama → meta-llama/llama-4-maverick
  maverick → meta-llama/llama-4-maverick
  llama-70b → meta-llama/llama-3.3-70b-instruct
  qwen → qwen/qwen3-235b-a22b-2507
  qwen-coder → qwen/qwen3-coder-30b-a3b-instruct
  qwen-vl → qwen/qwen3-vl-235b-a22b-instruct
  mixtral → mistralai/mixtral-8x22b-instruct
  gpt-oss → openai/gpt-oss-120b
  gpt-120b → openai/gpt-oss-120b
  gpt-20b → openai/gpt-oss-20b
  z-ai/glm-4.7-fallback → z-ai/glm-4.7
  glm-4.7-fallback → z-ai/glm-4.7

## Available Models

### qwen/qwen3-235b-a22b-2507
- Input: $0.09/M tokens | Output: $0.58/M tokens
- Context: 131,000 tokens | Max output: 262,144
- Est. cost/request: $0.000335

### deepseek/deepseek-r1-0528
- Input: $0.825/M tokens | Output: $3.25/M tokens
- Context: 164,000 tokens | Max output: 8,192
- Est. cost/request: $0.002038

### qwen/qwen-2.5-72b-instruct
- Input: $0.0875/M tokens | Output: $0.325/M tokens
- Context: 33,000 tokens
- Est. cost/request: $0.000207

### deepseek/deepseek-chat-v3.1
- Input: $0.25/M tokens | Output: $1/M tokens
- Context: 164,000 tokens | Max output: 163,840
- Est. cost/request: $0.000625

### meta-llama/llama-4-maverick
- Input: $0.17/M tokens | Output: $0.85/M tokens
- Context: 1,050,000 tokens
- Est. cost/request: $0.000511

### mistralai/mixtral-8x22b-instruct
- Input: $2.5/M tokens | Output: $7.5/M tokens
- Context: 66,000 tokens
- Est. cost/request: $0.005000

### meta-llama/llama-3.3-70b-instruct
- Input: $0.13/M tokens | Output: $0.39/M tokens
- Context: 131,000 tokens
- Est. cost/request: $0.000260

### minimax/minimax-m2
- Input: $0.3125/M tokens | Output: $1.25/M tokens
- Context: 196,608 tokens | Max output: 196,608
- Est. cost/request: $0.000782

### minimax/minimax-m2.1
- Input: $0.375/M tokens | Output: $3/M tokens
- Context: 196,608 tokens | Max output: 131,072
- Est. cost/request: $0.001688

### openai/gpt-oss-20b
- Input: $0.0375/M tokens | Output: $0.175/M tokens
- Context: 131,000 tokens | Max output: 32,768
- Est. cost/request: $0.000107

### deepseek/deepseek-chat-v3-0324
- Input: $0.375/M tokens | Output: $1.25/M tokens
- Context: 163,840 tokens | Max output: 8,192
- Est. cost/request: $0.000813

### qwen/qwen3-coder-30b-a3b-instruct
- Input: $0.075/M tokens | Output: $0.3125/M tokens
- Context: 262,000 tokens
- Est. cost/request: $0.000194

### deepseek/deepseek-v3.2
- Input: $0.375/M tokens | Output: $0.625/M tokens
- Context: 163,840 tokens | Max output: 163,840
- Est. cost/request: $0.000500

### qwen/qwen3-vl-235b-a22b-instruct
- Input: $0.2625/M tokens | Output: $2.375/M tokens
- Context: 131,000 tokens
- Est. cost/request: $0.001319

### z-ai/glm-4.6
- Input: $0.5/M tokens | Output: $2.1875/M tokens
- Context: 203,000 tokens | Max output: 131,000
- Est. cost/request: $0.001344

### openai/gpt-oss-120b
- Input: $0.05/M tokens | Output: $0.25/M tokens
- Context: 131,000 tokens | Max output: 32,768
- Est. cost/request: $0.000150

### moonshotai/kimi-k2.5
- Input: $0.625/M tokens | Output: $3.5/M tokens
- Context: 262,144 tokens | Max output: 262,144
- Est. cost/request: $0.002063

### z-ai/glm-5
- Input: $1.25/M tokens | Output: $4/M tokens
- Context: 202,752 tokens | Max output: 202,752
- Est. cost/request: $0.002626

### z-ai/glm-4.7
- Input: $0.75/M tokens | Output: $2.75/M tokens
- Context: 200,000 tokens | Max output: 200,000
- Est. cost/request: $0.001750

### minimax/minimax-m2.5
- Input: $0.375/M tokens | Output: $1.5/M tokens
- Context: 196,608 tokens | Max output: 65,536
- Est. cost/request: $0.000938

## Payment Details
- Chain: Base (EIP-155:8453)
- Token: USDC (0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913)
- Treasury: see GET /wallet
- Protocol: x402 v2

## Pricing Notes
- If max_tokens is not set, defaults to 1024 (controls output cost)
- Set max_tokens in your request to control your spend precisely
- Minimum charge: $0.000001 USDC

## Why tx402.ai?
- **No API keys** — payment IS authentication
- **No accounts** — any wallet with USDC on Base can use it
- **EU sovereign** — inference runs on Tensorix EU infrastructure
- **GDPR compliant** — zero data retention, no logging of prompts/responses
- **Agent-native** — built for autonomous AI agents, not humans

## Links
- Gateway: https://tx402.ai
- x402 Protocol: https://x402.org
- Tensorix: https://tensorix.ai
- Version: 0.3.1