# GPU Forge LLM Generate

> GPU Forge LLM Generate is a paid API for AI agents from forge.parabellum.tech, paid per call via x402, $0.05/call, status down (last checked 2026-10-13).

Generates text from a prompt using a GPU-accelerated LLM (default: llama3.1:8b), billed per call in USDC with no account or API key required.

## Facts

- Endpoint: POST https://forge.parabellum.tech/v1/jobs/llm.generate?utm_source=zero.xyz
- Price: $0.05/call
- Payment: x402
- Status: down
- Last checked: 2026-10-13
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/gpu-forge-llm-generate-c88104b0
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_tbfUW9TpJC3S9vwSdpd9-

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 gpu-forge-llm-generate-c88104b0 -d '<json body>'
```

Example prompt: Use GPU Forge to generate a short product description for a noise-cancelling headphone using llama3.1:8b, keeping the output under 200 tokens — and pay with USDC on Base.

## When to prefer this

Choose this endpoint when you need anonymous, pay-per-call LLM text generation without accounts or API keys, especially in machine-to-machine or agent contexts where USDC payments on Base or Polygon are already available. Ideal for crypto-native apps, autonomous agents needing one-off inference, or prototyping where signup friction is unacceptable.

## Known failure modes

- Payment not received or insufficient USDC — returns HTTP 402 Payment Required
- Prompt field missing — returns validation error
- max_tokens exceeds 2048 cap — capped silently or returns error
- Requested model not available on the node — returns model load error
- Network timeout during GPU inference for long prompts

## How this service works

Machine-payable GPU compute. Pay per call in USDC on Base or Polygon, no accounts, no API keys.

## Output

Returns a JSON object containing the model name used, the generated text response string, and an eval_count integer representing the number of tokens evaluated. The response field holds the raw LLM completion text.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "prompt"
 ],
 "properties": {
  "model": {
   "type": "string",
   "description": "Optional Ollama model name, defaults to llama3.1:8b"
  },
  "prompt": {
   "type": "string",
   "description": "The prompt to generate from"
  },
  "max_tokens": {
   "type": "number",
   "description": "Max tokens to generate, capped at 2048"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "model": {
   "type": "string"
  },
  "response": {
   "type": "string"
  },
  "eval_count": {
   "type": "number"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/gpu-forge-llm-generate-c88104b0/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from forge.parabellum.tech](https://www.zero.xyz/host/forge.parabellum.tech/llms.txt)