> ## Documentation Index
> Fetch the complete documentation index at: https://docs.uncensoredgpt.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# API Introduction

> Every UncensoredGPT model through one OpenAI-compatible API. Pay per token, no filters, no logs.

> [!NOTE]
>
> Prefer a web UI? [UncensoredGPT Chat](https://chat.uncensoredgpt.ai) runs on the same models — your API usage and your chat share the same platform.

The UncensoredGPT API gives you programmatic access to every model in the catalog through a single gateway. The endpoints follow the OpenAI API specification, so any library or framework that speaks that format works out of the box — just change the base URL and the API key.

## Base URL

```
https://api.uncensoredgpt.ai
```

All endpoints are prefixed with `/v1`. For example: `POST https://api.uncensoredgpt.ai/v1/chat/completions`.

You can also see the base URL for your account on the **Overview** page of the [UncensoredGPT Console](https://console.uncensoredgpt.ai).

## What you can do

* **Chat** — send a conversation, get a response. Streaming, tool calling, and vision models are supported.
* **Responses** — the newer stateful generation API, with tool use and retrieval of past responses.
* **Completions** — the legacy text-completion format.
* **Embeddings** — turn text into vectors for search and retrieval.
* **Images** — generate and edit images.
* **Batch API** — submit large workloads asynchronously at lower cost.
* **Files** — upload and manage files, for example Batch input files.
* **Messages** — call models in the Anthropic Messages format if your code already targets it.
* **Coding agents** — [OpenCode](https://docs.uncensoredgpt.ai/api-reference/agents/opencode.md), [Pi](https://docs.uncensoredgpt.ai/api-reference/agents/pi.md), [Claude Code](https://docs.uncensoredgpt.ai/api-reference/agents/claude-code.md), and any OpenAI-compatible tool connect in about a minute — see [Use with coding agents](https://docs.uncensoredgpt.ai/api-reference/agents.md).

## Make your first request

```bash theme={null}
curl -X POST https://api.uncensoredgpt.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $UNCENSOREDGPT_API_KEY" \
  -d '{
    "model": "uncensoredgpt/qwen3.8-27b-uncensored",
    "messages": [
      {"role": "user", "content": "Hello, what can you do?"}
    ]
  }'
```

## Next steps

> [card] [Authentication](https://docs.uncensoredgpt.ai/api-reference/authentication.md)
>
> Create an API key and authenticate your requests.

> [card] [Models](https://docs.uncensoredgpt.ai/api-reference/models.md)
>
> See the available models and how to list them from your code.

> [card] [Rate limits](https://docs.uncensoredgpt.ai/api-reference/rate-limits.md)
>
> Request and token limits, and how to handle them.

> [card] [Error codes](https://docs.uncensoredgpt.ai/api-reference/error-codes.md)
>
> What each error means and how to recover from it.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.