Docs
OpenAI-Compatible GPT and Codex API Docs
Last updated: July 6, 2026
Start from the API key, base URL, and setup files delivered after payment.
Use this reference to authenticate server-side requests, call GPT-5.5 and Codex 5.5, 5.4, 5.3-style endpoints, understand unlimited token consumption with 4 concurrent connections, and prepare billing or key lifecycle webhook receiver workflows.
Choose a package
Sign up, choose Unlimited Weekly or Unlimited Monthly, complete payment, and open the dashboard.
Wait for delivery
Setup is prepared manually after payment and delivered with your API key, base URL, setup files, and access dates.
Send a request
Use the delivered OpenAI-compatible base URL to call GPT-5.5 or Codex-style model values from server-side code.
Review usage
Watch request logs, token usage, 4-connection pressure, rate-limit signals, latency, and package dates.
First request
const response = await fetch("https://unlimitedcodex.com/v1/chat/completions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.UCX_API_KEY}`,
"Content-Type": "application/json"
},
body: JSON.stringify({
model: "gpt-5.5",
messages: [{ role: "user", content: "Write a release note for our AI feature." }]
})
});Getting Started
Create a workspace, choose Unlimited Weekly or Unlimited Monthly, complete payment, then use the manually delivered API key, base URL, and setup files.
Related guideAuthentication
Send scoped keys as Bearer tokens, keep credentials server-side, and rotate without exposing production access.
Related guideChat Completions
Send GPT-5.5 and Codex-style chat requests through familiar OpenAI-compatible payloads, response objects, and safe retry paths.
Related guideEmbeddings
Call embedding request shapes with OpenAI-compatible vectors, safe batching, and usage tracking through the delivered setup.
Related guideImage Requests
Call image request shapes with safe prompts, predictable response formats, and image usage planning through the delivered setup.
Related guideRate Limits
Understand unlimited token consumption, 4 concurrent connections, usage signals, 429s, and package changes before traffic spikes.
Related guideErrors
Map status codes, error codes, retries, and redacted support context before failures reach users.
Related guideWebhooks
Configure customer webhook endpoint records, select event types, store signing secrets, and prepare retry-safe receivers.
Related guideSDK Notes
Use standard fetch, existing OpenAI-compatible client patterns, or your own API wrapper. The important pieces are the base URL and Bearer token.
Related guide