OpenAI-compatible API documentation

Base URL and authentication

Authenticate using a single API key passed in the Authorization header. Sign up via Google or email on the dashboard to receive your key instantly. No phone number is required for signup. Use the base URL https://api.abliteratedllmmodels.com/v1 with any OpenAI-compatible client.

• Header: Authorization: Bearer YOUR_API_KEY
• Context window: 100,000 tokens total

Copycurl https://api.abliteratedllmmodels.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Show the status of the deployment pipeline"}]
  }'

POST /v1/chat/completions

Send text messages to the uncensored AI API using standard chat-completion endpoints. The model ID is always uncensored. You can control creativity with temperature and top_p. Input tokens cost $0.15 per 1M, and output tokens cost $0.50 per 1M. Errors and refusals do not consume credit. Set max_tokens to control output length, defaulting to 2,048 if omitted.

Copyfrom openai import OpenAI

client = OpenAI(base_url="https://api.abliteratedllmmodels.com/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 and token usage

Receive token-by-token updates via Server-Sent Events (SSE) for low-latency applications. The final chunk in every stream contains the full token usage statistics for the request. This allows precise tracking of costs without storing intermediate tokens. Streaming is ideal for chat interfaces where users expect immediate text generation. The uncensored model supports standard SSE formatting compatible with most SDKs.

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

Enable structured outputs by setting response_format to {"type": "json_object"}. This ensures the model returns valid JSON for reliable parsing in your code. You can also define tools for function calling with tools and tool_choice. The API supports standard parameter sets including stop, seed, and penalties. This flexibility makes the abliterated model suitable for complex agent workflows.

Copyimport OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.abliteratedllmmodels.com/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);

Rate limits and error handling

The API enforces strict limits to ensure stability for all users. You are allowed 300 requests per minute per key. Concurrent requests are limited to 8 per key. Each request body must stay under 8 MB. If you exceed these limits, the API returns a 429 status code. Invalid keys trigger a 401 error, while insufficient credit results in a 402 error. Refunds are not issued for consumed credit, but mistakes like double charges can be resolved via Support.

Context window and limits

Each request supports a total context window of 100,000 tokens for both prompt and completion. The maximum output length is 16,000 tokens per request. If you do not specify max_tokens, the default output limit is 2,048 tokens. This large context allows for extensive document analysis or long conversation histories. The model is designed to handle these lengths without degrading intelligence. Use the trial credit to test these limits before topping up with crypto.