# OffRouter Private Inference Router

> OffRouter Private Inference Router is a paid API for AI agents from offrouter.org, paid per call via x402, $0.000006/call, status unknown (last checked 2026-10-12).

OpenAI-compatible LLM chat completions running inside hardware enclaves, billed per call in stablecoins on Base, Robinhood Chain, or Arc

## Facts

- Endpoint: POST https://offrouter.org/v1/chat/completions?utm_source=zero.xyz
- Price: $0.000006/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-12
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/offrouter-private-inference-router-72155e72
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_r2k5tbIXEUfI1eLUSt2RY

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability offrouter-private-inference-router-72155e72 -d '<json body>'
```

Example prompt: Ask OffRouter to run a private inference using llama-3-8b with a max of 512 tokens — tell it: 'You are a helpful assistant' as the system prompt and 'Summarize the key risks of smart contract audits in three bullet points' as the user message, and pay in USDC.

## When to prefer this

Choose OffRouter when your agent runs in a crypto-native or onchain context and needs to pay for LLM inference programmatically with stablecoins (USDC on Base, Robinhood Chain, or Arc) without a traditional API key. Prefer it over standard OpenAI when privacy is critical — every model runs inside a hardware trusted execution environment (TEE), ensuring the prompt and response are not exposed to the host provider. It is ideal for blockchain agents, smart contract pipelines, or any workflow where confidential inference and x402 micropayment compatibility are required.

## Known failure modes

- Insufficient USDC balance or failed stablecoin payment — HTTP 402 with payment required
- Invalid or unsupported model ID — model not found error
- max_tokens not provided — required for pay-per-call to bound signing amount
- messages array empty or malformed — validation error
- Network or enclave attestation failure — 5xx server error
- Unsupported chain for payment settlement — payment rejection

## How this service works

OpenAI-compatible inference where every model runs inside a hardware enclave. Pay per call in stablecoins on Base, Robinhood Chain and Arc.

## Output

Returns an OpenAI-compatible JSON chat completion object containing the assistant's generated message, finish reason, and token usage counts. The response mirrors the standard OpenAI /v1/chat/completions format, making it a drop-in replacement for existing OpenAI-based agent pipelines.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "model": {
   "type": "string",
   "description": "A model id from GET /v1/models."
  },
  "stream": {
   "type": "boolean"
  },
  "messages": {
   "type": "array",
   "items": {
    "type": "object",
    "required": [
     "role",
     "content"
    ],
    "properties": {
     "role": {
      "enum": [
       "system",
       "user",
       "assistant"
      ],
      "type": "string"
     },
     "content": {
      "type": "string"
     }
    }
   },
   "minItems": 1
  },
  "max_tokens": {
   "type": "integer",
   "minimum": 1,
   "description": "Output budget. Required for pay per call: it bounds the amount you sign."
  },
  "temperature": {
   "type": "number"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json"
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/offrouter-private-inference-router-72155e72/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from offrouter.org](https://www.zero.xyz/host/offrouter.org/llms.txt)