Sonnet-level AI for ⅕ the cost.

One-line drop-in

Already using Anthropic or OpenAI? Add a baseURL override and an MWS API key. Your existing tool calls, streaming, and message shapes work unchanged.

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  baseURL: "https://api.mws.run/v1",
  apiKey: process.env.MWS_API_KEY,
});

const response = await client.messages.create({
  model: "claude-sonnet-4-6",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello" }],
});

Models

Two profiles are live today: Sonnet-class (default) — a drop-in for claude-sonnet-*model strings at much lower cost — and Pro (Think) for extended reasoning with a 1M-token context window. Pass any Claude or OpenAI model name and we'll route to the right one.

Rate limits auto-scale

Your rate limit grows automatically as your account spends and as you upgrade plans. There's no support ticket for higher throughput, and creating extra API keys doesn't increase your capacity — every key under your account draws from the same pool. Use additional keys to separate environments and rotate credentials, not to scale.

Every response carries the six standard x-ratelimit-* headers (compatible with OpenAI SDK retry logic), so your client can self-regulate without guessing or burning a request to discover the ceiling.

Frequently asked