Does the model actually refuse or just pretend?
The model does not refuse lawful adult, fictional, security-research or controversial topics. It will generate explicit content, roleplay scenarios, and unconventional narratives without the typical safety filters that kill the vibe. There is one hard limit that always applies: requests involving sexual content with minors are refused. This ensures you stay on the safe side of legal standards while keeping the creative freedom intact.
Many gateways claim to be uncensored but still filter based on their own policies. Our engine runs directly on our servers, so you get predictable behavior. You can test the boundaries in your own app without worrying about sudden policy shifts.
Context window size for long roleplay sessions
The context window supports up to 100,000 tokens for both prompt and completion combined. This allows for extensive roleplay sessions, long document analysis, or complex code generation without losing track of earlier details. You can maintain a rich history of the conversation while still having plenty of room for the model to generate a substantial response.
- Max output is 16,000 tokens per request.
- Default output is 2,048 tokens if you do not specify a limit.
- Longer contexts require careful management of the prompt size.
This capacity ensures that your long-running applications remain coherent and do not suffer from memory loss as the conversation progresses.
Pricing stability and token counting accuracy
Pricing is transparent and based strictly on token usage. You pay $0.15 per 1 million input tokens and $0.50 per 1 million output tokens. Errors and refusals do not incur charges, so you only pay for successful completions. This model protects your budget when dealing with unpredictable traffic patterns.
Credit is prepaid and never expires. You can top up with crypto only, accepting USDT on TRC20 or USDC on Base. The minimum top-up is $20 and the maximum is $500. You receive bonus credit depending on the amount: +5% from $50, +10% from $100, and +15% from $200. There are no monthly fees or subscriptions to manage.
OpenAI compatibility for easy migration
The API is fully OpenAI-compatible. You can use the official OpenAI SDKs or any client that supports the standard endpoints. The base URL is HTTPS://api.uncensoredaiopensource.com/v1. You just need to update the base_url and provide your API key.
Send requests to POST /v1/chat/completions and list models with GET /v1/models. The model id to use is "uncensored". This means you do not need to rewrite your entire stack to switch to an uncensored engine. It is a drop-in replacement for your current LLM provider.
Privacy: Are prompts used for training?
Privacy is a core feature. An account requires only an email address, and no phone number is needed. Your prompts are not used for training the model. This ensures that your proprietary data, roleplay scenarios, or creative content remain private.
You can sign up with Google or use an email and password. The API key is shown immediately. There is no complex verification process to slow down your development. Your data stays yours, which is crucial for commercial applications where content sensitivity matters.
Support for function calling and tools
The API supports function calling with tools and tool_choice parameters. You can define custom functions and let the model decide when to call them. This enables complex workflows, data extraction, and structured outputs. You can also force JSON mode with the response_format parameter.
Other supported parameters include temperature, top_p, stop, seed, presence_penalty, and frequency_penalty. This gives you fine-grained control over the model's behavior. You can tune the randomness, control output length, and ensure consistent results across different runs.
Streaming support for real-time UX
Streaming is supported via Server-Sent Events (SSE). This allows you to display tokens as they are generated, improving the user experience for chat interfaces. Token usage is included in the last chunk, so you can track costs in real-time.
This feature is essential for applications where latency matters. Users see the response forming instantly, rather than waiting for the entire completion. It makes the interaction feel more natural and responsive.
Rate limits for high-traffic apps
Rate limits are set to 300 requests per minute per key and 8 concurrent requests per key. The request body is limited to 8 MB. These limits are designed to ensure stability for high-traffic applications. If you need more concurrency, you can manage multiple keys or adjust your request frequency.
Each account has one active key. A new key replaces the old one, ensuring you always have a valid credential. This simplicity reduces administrative overhead while maintaining strict control over usage.
Crypto payment flexibility
Payments are crypto-only, accepting USDT on TRC20 or USDC on Base. You can top up with any whole amount from $20 to $500. This appeals to developers who prefer decentralized payments and want to avoid traditional banking delays.
The trial credit is $0.50 for new accounts, valid for 7 days. This allows you to test the API thoroughly before committing to a larger top-up. Bonus credit is added automatically based on your top-up amount.
