Does most unhinged AI charge for failed requests?

No, errors and refusals are free. You only pay for tokens successfully processed. If a request fails due to invalid parameters or network issues, no tokens are deducted from your prepaid balance. This ensures you only pay for working inference.

How do I top up my account?

Top up using crypto only: USDT on TRC20 or USDC on Base. Minimum top-up is $20, maximum is $500. Bonus credit applies at higher tiers. Credit never expires. Use the dashboard to initiate a payment and receive a wallet address instantly.

Is the uncensored model trained on my prompts?

No, your prompts are not used for training. The model is a static open-weight instance. Your data remains private and is not repurposed for future model updates. You retain full control over your context without contributing to a shared dataset.

API Documentation | MostUnhingedAI

Authentication and keys

Obtain your API key immediately after signup. The platform uses a single active key per account, issued via Google or email login without phone verification. Pass this key in the Authorization header as a Bearer token. Each request validates credentials before processing tokens. Lost keys generate a replacement that invalidates the previous one instantly.

curl https://api.mostunhingedai.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Send my invoice for 50k tokens"}]
  }'

Keep your key secure. The header format follows standard conventions used by major SDKs. No special prefixes or scopes are required for basic text completion.

Chat completions endpoint

Send messages to POST /v1/chat/completions for inference. The base URL is https://api.mostunhingedai.com/v1. Requests accept standard message arrays with role and content fields. The model identifier is uncensored. You receive text responses without content refusals for lawful adult material. Errors like missing keys return 401 status codes.

from openai import OpenAI

client = OpenAI(base_url="https://api.mostunhingedai.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)

Context windows hold 100,000 tokens total. Output limits default to 2,048 tokens unless you specify a higher max_tokens value up to 16,000. Input and output tokens count against your prepaid balance.

Models endpoint

Query available models via GET /v1/models. This endpoint returns metadata about the single hosted model. It confirms the ID uncensored and its capabilities. No other models exist in this instance. Use this endpoint to verify connectivity before sending heavy payloads.

import OpenAI from "openai";

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

The response includes standard OpenAI-compatible fields. You can parse the model ID programmatically to ensure compatibility with your routing logic. This endpoint requires authentication but returns data instantly.

Parameters and settings

Control generation behavior with standard parameters. Use temperature for randomness and top_p for nucleus sampling. Set seed for reproducible outputs during debugging. The stop parameter halts generation at specific strings. Adjust presence_penalty and frequency_penalty to reduce repetition.

These settings affect the raw inference power directly. No hidden filters modify these values. You retain full control over the output distribution. Default values match common industry standards for stability.

Streaming responses

Enable real-time token delivery by setting stream: true. The API returns Server-Sent Events (SSE). Each chunk contains partial text and usage statistics. The final chunk includes the total token count for the request. Parse the choices array in each delta object.

stream = 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)

Streaming reduces perceived latency for long responses. Buffer chunks in your client application. The SSE format ensures reliable delivery even over unstable connections. Token usage updates reflect actual consumption accurately.

Limits, errors, and context

Respect the rate limit of 300 requests per minute per key. Concurrency caps at 8 simultaneous requests. Request bodies must stay under 8 MB. A 429 status indicates you exceeded these limits. Wait and retry with exponential backoff.

Errors carry specific codes. 401 means your key is invalid. 402 signals insufficient prepaid credit. Refunds are not available for consumed tokens, but unused credit never expires. The 100,000 token context window includes both prompt and completion tokens. Plan your input size carefully to avoid truncation.

API Documentation