---
title: "Paid: LLM Inference Proxy"
method: POST
path: "/api/x402/llm-proxy"
---

# Paid: LLM Inference Proxy

`POST /api/x402/llm-proxy`

## Request

- Base URL: `https://three.ws`
- URL: `https://three.ws/api/x402/llm-proxy`
- Auth: none required

three.ws LLM Inference Proxy — pay per completion with no API key required. Runs one-shot text prompts through the platform's free-first provider chain. Response includes measured latency, token counts, and the provider actually used. Ideal for latency benchmarking, agent pipelines, and one-off completions. Model aliases: "fast" (sub-second) · "smart" (quality backstop). Price: $0.005 USDC per completion on Base or Solana.

## Request body

- object
  - `model` string
  - `prompt` string, required
  - `max_tokens` integer

## Response `200`

three.ws LLM Inference Proxy result JSON

## Other responses

- `400`: Validation error
- `402`: Payment Required (x402)

---

[API](https://skmtc.dev/three/apis/three-ws-api.md) · [All operations](https://skmtc.dev/three/apis/three-ws-api/llms.txt) · [OpenAPI document](https://skmtc.dev/three/apis/three-ws-api/revisions/dbc5e3a81b79?raw)