# Claude Opus via x402 Gateway

> Claude Opus via x402 Gateway is a paid API for AI agents from x402-gateway-production.up.railway.app, paid per call via x402, $0.090000/call, status unknown (last checked 2026-09-13, last successful call 2026-04-19).

Runs a prompt through Anthropic's Claude Opus (most powerful model) and returns a high-quality completion, billed per-call via x402 micropayment

## Facts

- Endpoint: POST https://x402-gateway-production.up.railway.app/api/llm/claude-opus
- Price: $0.090000/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-13
- Last successful call: 2026-04-19
- Success rate: 100% of calls made through Zero
- Activations on Zero: 15
- Tags: x402
- Canonical page: https://www.zero.xyz/c/x402-gateway-production-up-railway-app-06b69221
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_fFxjNwZ_9GOtDQISGlWmD

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 x402-gateway-production-up-railway-app-06b69221 -d '<json body>'
```

Example prompt: Ask Claude Opus to analyze the trade-offs between microservices and monolithic architectures for a fintech startup handling high transaction volumes — I need a thorough, nuanced breakdown with specific recommendations.

## When to prefer this

Choose this endpoint when the task demands maximum reasoning depth, extended thinking, nuanced analysis, or complex multi-step coding — situations where cheaper/faster models fall short. Prefer Claude Opus over Claude Sonnet or Haiku siblings when quality matters more than cost or latency. Use when the user explicitly requests the most powerful Claude model or when prior attempts with lighter models produced unsatisfactory results.

## Known failure modes

- Insufficient USDC balance or failed x402 payment — HTTP 402 with payment required error
- Prompt exceeds context window limit — HTTP 400 with token count error
- Upstream Anthropic API unavailable — HTTP 502 or 503 from gateway
- Malformed messages array — HTTP 400 with schema validation error
- Rate limiting on gateway — HTTP 429 with retry-after header

## How this service works

Anthropic's most powerful model — best for complex analysis, coding, and extended thinking

## Output

Returns a JSON object containing the model's text completion, the model version used (e.g. anthropic/claude-4.6-opus), and detailed usage stats including total tokens, prompt tokens, completion tokens, and itemized cost breakdown in USD. The main answer is in the 'content' field.

## Example request

```json
{
 "messages": [
  {
   "role": "user",
   "content": "Explain the concept of recursion in programming with a simple example."
  }
 ]
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/x402-gateway-production-up-railway-app-06b69221/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from x402-gateway-production.up.railway.app](https://www.zero.xyz/host/x402-gateway-production.up.railway.app/llms.txt)