[OI] SDK with a custom base URL

The whole migration is one constructor argument. Keep your retries, streaming and error handling; swap the base URL and the model id.

Base URL

https://uncensoredai.dev/v1

Python

Pass base_url to the client constructor and use a catalog model id.

python

import os
from openai import [OI]

client = [OI](
    base_url="https://uncensoredai.dev/v1",
    api_key=os.environ["UNCENSOREDAI_API_KEY"],
)

stream = client.chat.completions.create(
    model="deepseek-v4.1-flash",
    messages=[{"role": "user", "content": "Hello"}],
    stream=True,
)
for chunk in stream:
    delta = chunk.choices[0].delta.content if chunk.choices else None
    if delta:
        print(delta, end="")

Node / TypeScript

Identical with the npm package.

typescript

import [OI] from "openai";

const client = new [OI]({
  baseURL: "https://uncensoredai.dev/v1",
  apiKey: process.env.UNCENSOREDAI_API_KEY,
});

cURL

Raw HTTP works too — useful for a quick smoke test.

curl

curl https://uncensoredai.dev/v1/chat/completions \
  -H "Authorization: Bearer sk-ua-XXXX-XXXX-XXXX-XXXX" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4.1-flash","messages":[{"role":"user","content":"Hello"}]}'

FAQ

Is it really the same SDK?+

Yes. UncensoredAI implements the [OI] chat-completions and models endpoints, including SSE streaming and the [OI] error envelope.

Where do I find valid model ids?+

The model catalog and GET /v1/models. Ids are case-sensitive.

More integrations

Generate access key All integrations