# x402engine HunyuanVideo3 (hy3) LLM Inference

> x402engine HunyuanVideo3 (hy3) LLM Inference is a paid API for AI agents from x402engine.app, paid per call via x402, $0.002335/call, status unknown (last checked 2026-10-24).

Runs inference against Tencent's HunyuanVideo3 (hy3) large language model via a pay-per-call HTTP 402 micropayment API

## Facts

- Endpoint: POST https://x402engine.app/api/llm/hy3?utm_source=zero.xyz
- Price: $0.002335/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-24
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/x402engine-hunyuanvideo3-hy3-llm-inference-cc5228ec
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_ibMK7g_AlCD-Z7R4uAcc9

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 x402engine-hunyuanvideo3-hy3-llm-inference-cc5228ec -d '<json body>'
```

Example prompt: Send these messages to the Tencent hy3 model — system: 'You are a helpful assistant', user: 'Explain quantum entanglement in simple terms' — and stream back the response, paying per call with USDC on Base.

## When to prefer this

Choose this endpoint when you specifically want to run inference on Tencent's hy3 (HunyuanVideo3) language model via a trustless, pay-per-call micropayment model using USDC or USDm — ideal for agents that need no subscription, no API key management, and per-call billing on a capable Chinese-origin LLM with tool-calling and streaming support.

## Known failure modes

- Payment not settled — HTTP 402 returned if USDC/USDm payment is not included or insufficient
- Invalid messages array — missing required 'messages' field or malformed message objects returns 400
- Unsupported parameters — model may not support all sampling params (top_k, seed, etc.); check /docs/llm-capabilities.md
- Stream timeout — SSE stream may drop if network is unstable after payment settlement
- Tool call format errors — improperly formatted tool definitions or missing tool_call_id in tool responses

## How this service works

x402engine — Tencent's efficient reasoning model with a 262k-token context window

## Output

Returns an OpenAI-compatible JSON response containing the model name ('tencent/hy3'), a choices array with the assistant message (role, content, finish_reason, index), and a top-level content field with the reply text. When streaming is enabled, returns live OpenAI-compatible SSE events after payment settlement.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "required": [
  "input"
 ],
 "properties": {
  "input": {
   "type": "object",
   "required": [
    "type",
    "bodyType",
    "body",
    "method"
   ],
   "properties": {
    "body": {
     "type": "object",
     "required": [
      "messages"
     ],
     "properties": {
      "seed": {
       "type": "integer",
       "description": "Model-supported seed; see /docs/llm-capabilities.md."
      },
      "stop": {
       "oneOf": [
        {
         "type": "string"
        },
        {
         "type": "array",
         "items": {
          "type": "string"
         },
         "maxItems": 4
        }
       ],
       "description": "Model-supported stop; see /docs/llm-capabilities.md."
      },
      "min_p": {
       "type": "number",
       "description": "Model-supported min_p; see /docs/llm-capabilities.md."
      },
      "tools": {
       "type": "array",
       "items": {
        "type": "object",
        "required": [
         "function"
        ],
        "properties": {
         "type": {
          "enum": [
           "function"
          ],
          "type": "string"
         },
         "function": {
          "type": "object",
          "properties": {
           "name": {
            "type": "string"
           },
           "strict": {
            "type": "boolean"
           },
           "parameters": {
            "type": "object"
           },
           "description": {
            "type": "string"
           }
          }
         }
        }
       },
       "description": "Client-executed function definitions; return results using role=tool and tool_call_id."
      },
      "top_k": {
       "type": "integer",
       "description": "Model-supported top_k; see /docs/llm-capabilities.md."
      },
      "top_p": {
       "type": "number",
       "description": "Model-supported top_p; see /docs/llm-capabilities.md."
      },
      "stream": {
       "type": "boolean",
       "description": "Return live OpenAI-compatible SSE after payment settlement."
      },
      "messages": {
       "type": "array",
       "items": {
        "type": "object",
        "anyOf": [
         {
          "required": [
           "content"
          ]
         },
         {
          "required": [
           "tool_calls"
          ]
         }
        ],
        "required": [
         "role"
        ],
        "properties": {
         "name": {
          "type": "string"
… (truncated)
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "model": "tencent/hy3",
  "choices": [
   {
    "index": 0,
    "message": {
     "role": "assistant",
     "content": "Hello!"
    },
    "finish_reason": "stop"
   }
  ],
  "content": "Hello!"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/x402engine-hunyuanvideo3-hy3-llm-inference-cc5228ec/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from x402engine.app](https://www.zero.xyz/host/x402engine.app/llms.txt)