Azure for compliance, ProxyLLM for the rest.
Azure OpenAI lives inside enterprise cloud policy. ProxyLLM covers the standard OpenAI side: your own key passes through one endpoint, and OpenAI-bound work can run on Codex Hosted against a flat ChatGPT subscription instead of per-token pricing.
$129/month SaaS. Bring your own model keys. No inference markup.
Three steps to connect.
Know the Azure boundary
Azure OpenAI uses deployment-specific URLs and api-version parameters. ProxyLLM does not claim native Azure key storage today, so keep policy-bound deployments where they are.
Centralize the standard OpenAI side
Point OpenAI-compatible clients at https://api.proxyllm.ai/v1 with your own OpenAI or OpenRouter key. Same request shape, one endpoint, a log line for every call.
Scope team access
Issue scoped sub-keys with budget caps so each app spends against a limit you set, not against a shared secret.
The standard OpenAI side, one endpoint.
Pass gpt-4o-mini and other standard OpenAI models through ProxyLLM when a request does not need an Azure-specific deployment.
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: "gpt-4o-mini",
messages: [{ role: "user", content: "Run the non-Azure fallback path." }],
});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.
Give OpenAI traffic a flat bill.
Codex Hosted runs OpenAI's Codex on ProxyLLM servers, signed in with your ChatGPT account. Azure-specific deployments stay on Azure; everything standard gets caps, logs, and no inference markup.