Grok without another client path.
Treat Grok as one more model behind the ProxyLLM endpoint. Use x-ai/grok-3 for tone and realtime-flavored product work while your application keeps one SDK.
$129/month SaaS. Bring your own model keys. No inference markup.
Three steps to connect.
Use an OpenRouter-backed path
Grok models are reachable through OpenRouter where your account has access. Native xAI key support is a future direct-provider step, so Grok passes through on your own OpenRouter key today.
Keep the OpenAI client
Set baseURL to https://api.proxyllm.ai/v1 and use your ProxyLLM key. Your application keeps the same chat completions shape.
Track real usage
Put Grok beside OpenAI, Claude, Gemini, and Llama in the same request logs so cost and latency per model family stay visible.
Same SDK, Grok included.
Use the x-ai/ model prefix for OpenRouter-backed Grok access where your account allows it.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.proxyllm.ai/v1",
apiKey: "pk_live_...",
});
const r = await client.chat.completions.create({
model: "x-ai/grok-3",
messages: [{ role: "user", content: "Draft a punchy product note." }],
});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.
Compare Grok in production.
ProxyLLM logs latency, tokens, and cost beside your other model families, so you decide with request data instead of vibes.