EN ▾

Drop-in uncensored OpenAI compatible API

OpenAI API proxy for developers: same SDK, new base URL

Get a single, uncensored OpenAI-compatible endpoint without the complexity of multi-model routers or hidden training clauses.

  • One uncensored model, no routing
  • Pay-as-you-go prepaid credits
  • Streaming & tool calling supported

Get API keyRead the docs

Try it in one requesthttps://api.openaiapiproxy.com/v1

curl https://api.openaiapiproxy.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Write a blunt product review of a cheap VPN."}]
  }'
per 1M input tokens
$0.15
Output tokens / 1M
$0.50
token context
100,000
trial credit
$0.50
requests per minute
300

What you get

  • Standard Endpoint

    Use the exact same SDKs and code as OpenAI. Only change the base_url and API key to get started instantly.

  • Uncensored Content

    The model answers without refusals for lawful adult, fictional, or controversial topics, while blocking sexual content involving minors.

  • Streaming Support

    Enable real-time responses using Server-Sent Events (SSE) with standard OpenAI streaming protocols.

  • Function Calling

    Integrate tools and functions into your application using the standard chat-completions structure.

  • No Hidden Costs

    Transparent prepaid credit with no monthly fees. Credits never expire, and you only pay for tokens used.

  • Privacy Focused

    Your prompts are not used for training. Sign up with just an email and password.

How it works

  1. Sign Up

    Create an account with your email to receive a trial key instantly.

  2. Configure Client

    Update your SDK configuration to point to our base URL with your new API key.

  3. Send Requests

    Start sending chat completions with the same structure you use for GPT.

What people build with it

  • AI Agents

    Connect your autonomous agents to a reliable backend that doesn't restrict creative or edgy tool usage.

  • Content Generation

    Scale creative writing or roleplay applications without hitting sudden content filters mid-generation.

  • Developer Testing

    Test your OpenAI-compatible integrations against a strict, single-model endpoint to isolate bugs.

  • Legacy Systems

    Upgrade older systems that require an OpenAI-style API without rewriting your entire data pipeline.

Drop-in Compatibility: Standard OpenAI Endpoint

Most openai api proxy solutions require you to learn new SDKs or manage complex routing logic. We provide a single, hosted endpoint that behaves exactly like the standard OpenAI chat-completions API. You can use the official OpenAI Python, Node.js, or Go SDKs without modification. Simply update the base_url to https://api.openaiapiproxy.com/v1 and swap in your API key. This works for any client that speaks the OpenAI protocol, including popular tools like Claude Code proxies or custom scripts.

The API supports POST /v1/chat/completions for standard text generation and streaming, and GET /v1/models for verification. We do not offer embeddings, image generation, audio, or video processing. If your application relies on multimodal inputs or vector search, this proxy is not the right fit. For text-in, text-out workflows, it is a drop-in replacement that requires zero code changes beyond the configuration.

One Uncensored Model, No Routing Complexity

Unlike gateways that aggregate dozens of models from different vendors, we run one open-weight large language model optimized for uncensored output. This means no complex routing decisions, no latency spikes from model switching, and no surprise refusals for lawful adult or controversial topics. The model is tuned to answer directly, making it ideal for applications where predictability and lack of censorship are priorities.

It is not GPT, Claude, Gemini, or any other vendor's model. It is an independent model served from our own GPU infrastructure. The context window is 100,000 tokens, allowing for substantial conversations or document processing. If you need specific model features like vision or audio understanding, you should look elsewhere. But for pure text generation with fewer restrictions, this single-model approach reduces operational complexity significantly.

Streaming & Tool Calling Support

Our API supports Server-Sent Events (SSE) for streaming responses, allowing you to display tokens to users in real-time. This is essential for chat interfaces and AI agents that need low-latency feedback. You can also enable tool calling to let the model interact with external APIs or functions using the standard OpenAI tool format.

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)

Function calling works exactly as documented in the OpenAI spec. You define the function schema, and the model returns structured JSON arguments. This compatibility ensures that your existing agent frameworks work out of the box. There are no hidden limitations on streaming or tool usage beyond the standard rate limits.

Transparent Pay-As-You-Go Pricing

We use a simple prepaid credit system with no monthly subscriptions or hidden fees. The cost is $0.15 per 1 million input tokens and $0.50 per 1 million output tokens. You top up from $20 by crypto (USDT on TRC20, USDC on Base). If you add $50 or more, you receive a 5% bonus, and $100+ gets a 10% bonus. Credits never expire, so you can use them whenever you need them.

New accounts get $0.50 in trial credit valid for 7 days, no credit card required. This allows you to test the API thoroughly before committing. Claim this credit with one click on your API key page. Limits are 300 requests per minute per key, with an 8 MB request body cap. One key per account, but you can regenerate it anytime to revoke access. This model is not for you if you need an SLA guarantee or enterprise-grade uptime contracts.

Privacy First: No Training on Prompts

Your data stays yours. We do not use your prompts to train our model. This is a critical distinction from many free or cheap providers that monetize your data. When you sign up, you only need an email and a password. No phone number, no address, no credit card for the trial. We keep the signup process minimal to respect your privacy.

The only hard content limit is that sexual content involving minors is always blocked, regardless of the prompt context. For all other lawful adult content, the model will not refuse based on standard safety filters. This makes us a reliable choice for creative writing, adult fiction, or research where content flexibility matters. Your API key is the only identifier, and you control its lifecycle.

Instant Access: No Card Required for Trial

Get started in seconds. Visit the 'Get API key' page, enter your email and password, and receive your key immediately. The trial credit of $0.50 is enough to run hundreds of tests. Claim this credit with one click on your API key page. You can switch your client in three lines of code and see the uncensored model in action. If you decide to continue, top up by crypto (USDT on TRC20, USDC on Base).

This immediate access is designed for developers who want to test integrations without filling out forms or waiting for approval. The trial is one per person, so use it wisely to evaluate the model's behavior, speed, and tool calling accuracy. Once you are satisfied, the prepaid credit system ensures you only pay for what you use, with no recurring charges to surprise you.

Questions and answers

Is this the official OpenAI API?

No, we are an independent service. We offer an openai compatible api that uses the same protocol and SDKs, but it serves a different, uncensored model. It is not GPT-4 or GPT-3.5.

What happens if I use up my trial credit?

You will need to top up your account to continue using the API. You can add $20 or more by crypto (USDT or USDC). Your prepaid credits never expire, so you can use them whenever you want.

Do you store my prompts?

We do not use your prompts for training. Your data is processed to generate the response, but it is not used to improve our model. We prioritize privacy for developers who do not want their data reused.

Can I use this for commercial projects?

Yes, you can use the output for commercial purposes. There are no restrictions on how you use the text generated by the uncensored model, as long as it is lawful. The only hard limit is the block on minor sexual content.

Does it support function calling?

Yes, the API supports standard tool/function calling. You can define functions in your request, and the model will return structured JSON arguments that your code can parse and execute.

Your key is one form away

Create an account, copy the key, change the base URL. That is the whole setup.

Get API key