# JarvisClaw Compute Sandbox Create

> JarvisClaw Compute Sandbox Create is a paid API for AI agents from api.jarvisclaw.ai, paid per call via x402, $0.045/call, status unknown (last checked 2026-09-15).

Creates a sandboxed AI compute session via OpenAI-compatible chat completions API with per-call USDC micropayments over x402

## Facts

- Endpoint: GET https://api.jarvisclaw.ai/v1/marketplace/compute/sandbox/create
- Price: $0.045/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/jarvisclaw-compute-sandbox-create-0e6b6bc7
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_GsHsPbWeWatU6tfRCbbu-

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 jarvisclaw-compute-sandbox-create-0e6b6bc7
```

Example prompt: Use JarvisClaw to send these messages to gpt-4.1-nano with a temperature of 0.7 and max 512 tokens, streaming the response: [{"role":"user","content":"Explain quantum entanglement in simple terms"}]

## When to prefer this

Choose this endpoint when you need on-demand LLM inference without a monthly subscription, want to pay per call in USDC via the x402 protocol, require OpenAI-compatible API compatibility for drop-in use, need smart routing across multiple model providers (OpenAI, Anthropic), or want isolated sandbox compute sessions for each request.

## Known failure modes

- 402 Payment Required — x402 micropayment not attached or insufficient USDC balance
- 400 Bad Request — invalid model ID, malformed messages array, or out-of-range temperature/max_tokens
- 401 Unauthorized — missing or invalid API credentials
- 429 Too Many Requests — rate limit exceeded on the gateway
- 503 Service Unavailable — underlying model provider unavailable or sandbox creation failed
- 408 Timeout — model inference exceeded allowed compute time for sandbox tier

## How this service works

AI API Gateway with smart routing, pay per call via x402. OpenAI-compatible. Settled in USDC on Base & Solana.

## Output

A chat completion object (OpenAI-compatible format) containing the AI-generated response text, token usage statistics, and model metadata; if streaming is enabled, returns a stream of Server-Sent Events with delta tokens.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "query": {
   "type": "string",
   "description": "Request parameters"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/jarvisclaw-compute-sandbox-create-0e6b6bc7/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.jarvisclaw.ai](https://www.zero.xyz/host/api.jarvisclaw.ai/llms.txt)