Use OpenRouter access, with ProxyLLM controls.
OpenRouter is the model catalog. ProxyLLM is the control layer in front of it: app-level keys, request logs, and budget caps. OpenAI-bound work can run through Codex Hosted on your ChatGPT subscription instead.
$129/month SaaS. Bring your own model keys. No inference markup.
Three steps to connect.
Add your OpenRouter key
Paste the OpenRouter key in ProxyLLM once. ProxyLLM stores it encrypted and uses it when a model name has a provider prefix.
Keep one app endpoint
Point your app to https://api.proxyllm.ai/v1 with a ProxyLLM key instead of exposing OpenRouter directly to every caller.
Add logs and caps
OpenRouter supplies broad model access as the passthrough lane. ProxyLLM adds scoped app keys, request logs, and spend limits in front of it. Anything bound for OpenAI models can run on Codex Hosted instead.
Request provider-prefixed models.
Keep OpenRouter behind ProxyLLM and use the same OpenAI-compatible client.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.proxyllm.ai/v1",
apiKey: process.env.PROXYLLM_API_KEY,
});
await client.chat.completions.create({
model: "anthropic/claude-3.5-sonnet",
messages: [{ role: "user", content: "Review this design." }],
});Run your AI workloads on your ChatGPT subscription.
ProxyLLM runs OpenAI's Codex for you, signed in with your own ChatGPT account. Your apps call one OpenAI-compatible endpoint and the work bills to your flat plan instead of per-token API pricing.
Put governance in front of the router.
Give apps their own ProxyLLM keys while one encrypted OpenRouter key stays managed centrally.