Unfiltered LLM API: Call It in Five Minutes
An unfiltered LLM API is a hosted model you call over HTTP that answers without content refusals for lawful adult use. This quickstart gets you from zero to a streaming response in five minutes.
Updated


Key points
- The API is OpenAI-compatible: change the base URL and key, keep your code.
- Base URL is https://api.unfilteredaimodels.com/v1 and the model ID is uncensored.
- Streaming, tool calling and JSON mode work; text only, 100,000-token context.
- Pricing is $0.15 per 1M input and $0.50 per 1M output, prepaid, no subscription.
The short version
- Create an account and open the API key page. Claim the $0.50 trial credit if you want to test first; it needs no card.
- Point any OpenAI-compatible client at
https://api.unfilteredaimodels.com/v1. - Send the key as
Authorization: Bearer <key>and use modeluncensored. - Call
POST /v1/chat/completions, withstream: trueif you want tokens as they are produced.
That is the whole integration. If you already have code written for an OpenAI SDK, you change two strings and run it. The rest of this page shows each step with code you can paste, plus the details that save time later: limits, balance checks and common errors.
What you are actually calling
Be clear about the product before you build on it. The service hosts one open-weight model, run by an independent provider and tuned to answer without content refusals for lawful adult use. It is not the official API of any model vendor, and it is not a router that forwards your request elsewhere. One model, one model ID.
That has practical consequences. You do not pick between model sizes, and you cannot swap in another vendor's model by changing a string. In return, behaviour stays consistent from one call to the next, which is what you want when a script depends on it. The unfiltered AI model page explains the model side in more detail.
Two hard boundaries apply. Content sexualising minors is not allowed, and the API is text only: no images, audio, video, embeddings or fine-tuning.
First request with curl
export API_KEY="paste-your-key-here"
curl https://api.unfilteredaimodels.com/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "uncensored",
"messages": [
{"role": "system", "content": "You are a direct, concise assistant."},
{"role": "user", "content": "Give me three names for a fantasy tavern."}
],
"temperature": 0.9
}'
The reply follows the OpenAI schema: choices[0].message.content holds the text and usage reports token counts. The response headers include X-Request-Id, X-Balance-USD, X-RateLimit-Limit-Requests and X-RateLimit-Limit-Concurrency.
If you prefer working from a page that only uses curl and a token, the curl with token guide goes deeper on that style. The newer POST /v1/responses method is also available for clients that use it, and GET /v1/models lists the model.
Python and Node
Use the official OpenAI libraries; there is no special SDK to install.
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["API_KEY"],
base_url="https://api.unfilteredaimodels.com/v1",
)
resp = client.chat.completions.create(
model="uncensored",
messages=[{"role": "user", "content": "Summarise the plot of Hamlet in 3 lines."}],
)
print(resp.choices[0].message.content)
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.API_KEY,
baseURL: "https://api.unfilteredaimodels.com/v1",
});
const resp = await client.chat.completions.create({
model: "uncensored",
messages: [{ role: "user", content: "Summarise the plot of Hamlet in 3 lines." }],
});
console.log(resp.choices[0].message.content);
Frameworks that accept a custom OpenAI base URL, such as chat front ends and agent libraries, work the same way. Look for fields named base URL, API base or endpoint, and enter the model ID by hand if the tool does not fetch the model list.
Streaming
For chat interfaces you want tokens as they are produced. Set stream: true and read server-sent events. Usage is sent at the end of the stream.
stream = client.chat.completions.create(
model="uncensored",
messages=[{"role": "user", "content": "Write a four-line poem about rain."}],
stream=True,
)
for chunk in stream:
if chunk.choices and chunk.choices[0].delta.content:
print(chunk.choices[0].delta.content, end="", flush=True)
One detail to remember: a streamed request keeps its concurrency slot until the stream ends. A full key allows 8 concurrent requests, so a web app with many readers should close streams promptly and queue the rest.
Tool calling (tools, tool_choice, results sent back as role: "tool") and JSON mode (response_format: {"type":"json_object"}) are supported, as are temperature, top_p, stop, seed and the penalties.
JSON mode for structured output
When a script needs to parse the answer, ask for JSON instead of scraping prose. Set response_format and say in the prompt that you want JSON, with the keys you expect.
curl https://api.unfilteredaimodels.com/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "uncensored",
"response_format": {"type": "json_object"},
"messages": [{"role": "user", "content": "Return JSON with keys name and tagline for an imaginary bakery."}]
}'
Still wrap the parse in a try block. JSON mode makes valid output the normal case, but a reply cut off by a stop sequence or a very small max_tokens can end mid-object. Setting seed and a low temperature makes repeated runs more alike, which helps when you are debugging a prompt.
Check your balance
Billing is prepaid. Read the remaining credit three ways: the X-Balance-USD header on any response, GET /v1/balance, or the OpenAI-style GET /v1/dashboard/billing/credit_grants.
curl https://api.unfilteredaimodels.com/v1/balance \
-H "Authorization: Bearer $API_KEY"
Failed or refused requests are not charged. Prepaid credit never expires. Top-ups are any whole amount from $20 to $500, with a bonus of 5% from $50, 10% from $100 and 15% from $200. Pay by crypto (USDT on TRC20, USDC on Base). Full details sit on the pricing page.
Limits and the errors you will meet
A full key allows 8 concurrent requests and 300 requests per minute; a trial key allows 2 and 60. The context window is 100,000 tokens covering prompt plus completion, and the request body can be up to 8 MB. max_tokens is optional.
When you go over, you get HTTP 429. The error code rate_limited means too many requests per minute, and concurrency means too many at once. During short overload the API also answers 429 with a Retry-After: 5 header. A simple wrapper that sleeps for the header value and retries handles all three.
import time
from openai import RateLimitError
def ask(messages, tries=4):
for i in range(tries):
try:
return client.chat.completions.create(model="uncensored", messages=messages)
except RateLimitError:
time.sleep(5 * (i + 1))
raise RuntimeError("still limited")
Prompts are not used for training. For the complete parameter list, read the documentation.
More guides
Questions and answers
Your key is one form away
Create an account, copy the key, change the base URL. That is the whole setup.
Is this the official API of a specific model maker?
- No. It is one open-weight model run by an independent provider, tuned to answer without content refusals for lawful adult use. It is not any vendor's official API.
Do I need to change my OpenAI code much?
- Usually only the base URL, the API key and the model name, which is uncensored. The request and response format follow the OpenAI schema.
Can it generate images or embeddings?
- No. The API is text only. There is no image, audio or video generation, no embeddings and no fine-tuning.
How much does a request cost?
- Input is $0.15 per 1M tokens and output is $0.50 per 1M. A request with 1,000 input and 500 output tokens costs $0.00015 + $0.00025 = $0.0004.