Integration Guide: Chat Completions API

Base URL and Authentication

The abliterated API operates as a drop-in replacement for standard OpenAI endpoints. You route requests to https://api.aiabliterated.cc/v1 and authenticate using a Bearer token in the Authorization header. This setup works with any client that supports the OpenAI protocol.

• Base URL: https://api.aiabliterated.cc/v1
• Auth Header: Authorization: Bearer YOUR_API_KEY
• Model ID: uncensored

Authentication is strict. If you send an invalid key, the system returns a 401 error immediately.

The /v1/chat/completions Endpoint

Send your prompt data to the /v1/chat/completions endpoint to receive text responses. The system expects a JSON payload containing the model identifier and your message history. This is a text-only interface; do not expect image or audio generation.

The model processes input and output within a 100,000 token context window. If you do not specify a limit, the system defaults to generating 2,048 tokens per request. You can adjust this up to a hard maximum of 16,000 tokens.

Copyfrom openai import OpenAI

client = OpenAI(base_url="https://api.aiabliterated.cc/v1", api_key="YOUR_KEY")

resp = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Summarise this thread without softening it."}],
)
print(resp.choices[0].message.content)

Streaming Responses & Token Usage

Enable streaming by setting stream: true in your request payload. The API returns a Server-Sent Events (SSE) stream of partial tokens. This reduces perceived latency significantly for interactive applications.

Token usage data is not sent with every chunk. Instead, the final chunk of the stream contains the complete usage statistics for that request. This design keeps the stream lightweight while providing accurate billing data. Errors during streaming are handled by closing the connection with an error JSON object.

Copystream = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Tell the story in second person."}],
    stream=True,
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Function Calling and JSON Mode

The abliterated model supports structured outputs via the response_format parameter. Set this to {"type": "json_object"} to force strict JSON compliance. This is essential for reliable integration with backend systems.

Function calling is also available. Define tools in your request, and the model will output valid JSON matching your schema. The system does not validate your tool definitions, so ensure your schemas are correct before sending requests. This capability allows for complex agent architectures without external parsing layers.

Copyimport OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.aiabliterated.cc/v1", apiKey: process.env.API_KEY });

const resp = await client.chat.completions.create({
  model: "uncensored",
  messages: [{ role: "user", content: "Draft a villain monologue for my game." }],
});
console.log(resp.choices[0].message.content);

Parameters: Temperature, Top-P, and Seed

Control the randomness of your outputs using standard sampling parameters. The temperature value adjusts creativity, while top_p limits the nucleus sampling. Use seed for deterministic outputs when you need reproducible results for debugging.

Other available parameters include stop sequences, presence_penalty, and frequency_penalty. These allow fine-tuned control over token repetition and topic adherence. The uncensored model responds well to low temperature settings for code generation and high temperatures for creative writing.

Rate Limits and Constraints

The API enforces strict limits to ensure stability for all users. You are limited to 300 requests per minute per key. Concurrent requests are capped at 8 simultaneously. Each request body must not exceed 8 MB.

If you exceed these limits, the system returns a 429 Too Many Requests error. Billing is based on real token usage. Errors and refusals do not consume your prepaid credit. You can manage your key directly from the dashboard if you need to rotate credentials due to these constraints.

cURL

Copycurl https://api.aiabliterated.cc/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Summarize this 500 word technical document"}]
  }'