# LiteLLM

> Route to Kurrens from LiteLLM as an OpenAI-compatible endpoint.

Prefix the Kurrens model id with `openai/` and set `api_base`:

```python

response = litellm.completion(
    model="openai/deepseek-ai/DeepSeek-V4.1-Flash",
    api_base="https://api.kurrens.ai/v1",
    api_key="<KURRENS_API_KEY>",
    messages=[{"role": "user", "content": "Hello"}],
)
```

## LiteLLM proxy

```yaml
model_list:
  - model_name: deepseek-v4-flash
    litellm_params:
      model: openai/deepseek-ai/DeepSeek-V4.1-Flash
      api_base: https://api.kurrens.ai/v1
      api_key: os.environ/KURRENS_API_KEY
```