open-interpreter/docs/language-models/hosted-models/vllm.mdx

---
title: vLLM
---

To use Open Interpreter with vLLM, you will need to:

1. `pip install vllm`
2. Set the api_base flag:

<CodeGroup>

```bash Terminal
interpreter --api_base <https://your-hosted-vllm-server>
```

```python Python
from interpreter import interpreter

interpreter.llm.api_base = "<https://your-hosted-vllm-server>"
interpreter.chat()
```

</CodeGroup>

3. Set the `model` flag:

<CodeGroup>

```bash Terminal
interpreter --model vllm/<perplexity-model>
```

```python Python
from interpreter import interpreter

interpreter.llm.model = "vllm/<perplexity-model>"
interpreter.chat()
```

</CodeGroup>

# Supported Models

All models from VLLM should be supported