> ## 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.

# Rate Limits

> Request and token-per-minute limits, and how to handle 429 responses.

Rate limits are enforced per API key, in two dimensions:

| Dimension | What it measures | Default |
| - | - | - |
| Requests per minute (RPM) | How many requests you can make per minute | 60 |
| Tokens per minute (TPM) | How many tokens (input + output) you can process per minute | 1,000,000 |

When you exceed a limit, the API returns `429 Too Many Requests` until the window resets.

> [!NOTE]
>
> Limits can be adjusted per key when you create it in the Console's **Keys** page, and can be raised on request — contact [support](mailto:info@uncensoredgpt.ai).

## Check your limits

Rate-limit information is included in error responses, and your per-minute consumption is visible in real time in the Console's **Usage** page.

## Handle 429 responses gracefully

* Retry with exponential backoff and jitter, instead of retrying immediately.
* Respect the `retry-after` header when it is present.
* For batch workloads, use the [Batch API](https://docs.uncensoredgpt.ai/api-reference/endpoints/batches/create-batch.md), which processes large volumes asynchronously without hitting per-minute limits.

```python theme={null}
import time
from openai import OpenAI, RateLimitError

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.uncensoredgpt.ai/v1"
)

for attempt in range(5):
    try:
        response = client.chat.completions.create(
            model="uncensoredgpt/qwen3.8-27b-uncensored",
            messages=[{"role": "user", "content": "Hello!"}]
        )
        break
    except RateLimitError:
        time.sleep(2 ** attempt)
```


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