Pay-per-call LLM API: Llama 3.3 70B or gpt-oss-120b, no subscription

$0.01 per call · failed calls are never charged · no account or subscription: pay per call with a card key or USDC

Send a prompt, get an answer from a larger open model: Llama 3.3 70B Instruct (default) or OpenAI gpt-oss-120b, with an optional system prompt, up to 1,024 output tokens and adjustable temperature. Same call as /ai/chat with a stronger model. Use when the 8B model is not enough: reasoning, drafting and Q&A.

When to use it

Reasoning, drafting and Q&A when the 8B model is not enough
Pick llama-3.3-70b (default) or gpt-oss-120b with the model input
Call it through the OpenAI-compatible /v1/chat/completions by naming a 70b or gpt-oss model

Guide: Chat completions API, pay per call

Buy credits by card (from $1) Example result (JSON)

Inputs

Name	Type	Description
`prompt` (required)	string	User prompt, up to 8,000 characters
`model`	string	llama-3.3-70b (default) or gpt-oss-120b. One of: `llama-3.3-70b`, `gpt-oss-120b`.
`system`	string	Optional system prompt, up to 2,000 characters
`maxTokens`	integer	Maximum output tokens, 1-1024 (default 400)
`temperature`	number	0-1.5 (default 0.6)

Request

Over HTTP, with a card key from /credits:

curl -H "Authorization: Bearer adt_…" "https://tools.funwords.eu/ai/chat-large?prompt=Explain%20x402%20in%20one%20sentence."

Without a key, the same URL answers 402 Payment Required with x402 and MPP terms, so an agent with a wallet pays per call in USDC (Base, Polygon, Arbitrum or Solana over x402; Base or Tempo over MPP) and retries.

Example result

{
  "model": "@cf/meta/llama-3.3-70b-instruct-fp8-fast",
  "response": "x402 lets software pay for an HTTP request with a stablecoin by answering a 402 Payment Required challenge."
}

From an AI agent (MCP)

Add the MCP server to Claude, Cursor or any MCP client. This tool is ai_chat_large, and the first 25 successful calls are free:

https://tools.funwords.eu/mcp

Questions

How is it different from /ai/chat?

Same call and inputs (prompt, optional system prompt, up to 1,024 output tokens, temperature) with a stronger model, at its own price.

How do I pay?

Each successful call costs $0.01; failed calls are not charged. This tool is not in the free calls; pay with card credits (from $1; the key goes in an Authorization header or, for clients that can only fetch a URL, as api_key= in the URL) or per call in USDC over x402 or MPP. No account is needed.