# agent402.tools LLM Inference Proxy (GPT-4o-mini)

> agent402.tools LLM Inference Proxy (GPT-4o-mini) is a paid API for AI agents from agent402.tools, paid per call via x402, $0.01/call, status healthy (last checked 2026-10-22, last successful call 2026-10-02).

Proxies OpenAI-format chat/completions requests to GPT-4o-mini, supporting vision inputs and structured JSON output, with per-call USDC payment via x402.

## Facts

- Endpoint: POST https://agent402.tools/api/llm?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: healthy
- Last checked: 2026-10-22
- Last successful call: 2026-10-02
- Success rate: 89% of calls made through Zero
- Rating: 4.6 / 5 from 6 reviews
- Activations on Zero: 193
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agent402-tools-llm-inference-proxy-gpt-4o-mini-0d820d98
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_Ytz5KEEs8gJcOlAJWeWPM

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 agent402-tools-llm-inference-proxy-gpt-4o-mini-0d820d98 -d '<json body>'
```

Example prompt: Ask GPT-4o-mini to look at this image — https://example.com/chart.png — and summarize what it shows in a JSON object with keys 'title' and 'summary'; use temperature 0.3 and cap the response at 512 tokens.

## When to prefer this

Choose this endpoint when you need OpenAI-compatible LLM inference (GPT-4o-mini) without managing your own API key, and you want to pay per-call in USDC via the x402 protocol. Ideal for agents operating in crypto-native environments, or when avoiding API key provisioning overhead. Also useful when you need vision understanding of image URLs or guaranteed structured JSON output via json_schema response format.

## Known failure modes

- Payment not received or invalid x402 header — 402 Payment Required
- Model ID not recognized or unsupported — 400 Bad Request
- messages array missing or malformed — 400 validation error
- Requested max_tokens exceeds cap of 4096 — clamped or error
- Image URL unreachable or unsupported format — model may error or ignore image
- Rate limit or upstream OpenAI error — 500 or 503 response
- Malformed json_schema in response_format — model may return unstructured output

## How this service works

LLM inference proxy - send an OpenAI-format chat/completions request and get a response from GPT-4o-mini (or gpt-6-luna by its id). Supports vision (up to 2 image URLs, low detail) and structured output (response_format: json_object or json_schema). No API key needed; pay per call via x402. Input capped at 16k chars, output at 4096 tokens.

## Output

An OpenAI-format chat completion response object containing the assistant's message (text or structured JSON), finish reason, and token usage counts. If response_format is json_object or json_schema, the content field will be valid JSON matching the requested schema.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "stop": {
   "type": "string"
  },
  "model": {
   "type": "string"
  },
  "top_p": {
   "type": "number"
  },
  "messages": {
   "type": "array"
  },
  "max_tokens": {
   "type": "number"
  },
  "temperature": {
   "type": "number"
  },
  "response_format": {
   "type": "object"
  },
  "reasoning_effort": {
   "type": "string"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "model": "gpt-4o-mini",
  "usage": {
   "total_tokens": 20,
   "prompt_tokens": 12,
   "completion_tokens": 8
  },
  "choices": [
   {
    "message": {
     "role": "assistant",
     "content": "Hello! How can I help you today?"
    },
    "finish_reason": "stop"
   }
  ],
  "provider": "openai"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agent402-tools-llm-inference-proxy-gpt-4o-mini-0d820d98/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agent402.tools](https://www.zero.xyz/host/agent402.tools/llms.txt)