---
title: Arguments
---

<Card
  title="New: Streaming responses in Python"
  icon="arrow-up-right"
  href="/usage/python/streaming-response"
>
  Learn how to build Open Interpreter into your application.
</Card>

#### `messages`

This property holds a list of `messages` between the user and the interpreter.

You can use it to restore a conversation:

```python
interpreter.chat("Hi! Can you print hello world?")

print(interpreter.messages)

# This would output:

[
   {
      "role": "user",
      "message": "Hi! Can you print hello world?"
   },
   {
      "role": "assistant",
      "message": "Sure!"
   }
   {
      "role": "assistant",
      "language": "python",
      "code": "print('Hello, World!')",
      "output": "Hello, World!"
   }
]
```

You can use this to restore `interpreter` to a previous conversation.

```python
interpreter.messages = messages # A list that resembles the one above
```

---

#### `offline`

<Info>This replaced `interpreter.local` in the New Computer Update (`0.2.0`).</Info>

This boolean flag determines whether to enable or disable some offline features like [open procedures](https://open-procedures.replit.app/).

```python
interpreter.offline = True  # Check for updates, use procedures
interpreter.offline = False  # Don't check for updates, don't use procedures
```

Use this in conjunction with the `model` parameter to set your language model.

---

#### `auto_run`

Setting this flag to `True` allows Open Interpreter to automatically run the generated code without user confirmation.

```python
interpreter.auto_run = True  # Don't require user confirmation
interpreter.auto_run = False  # Require user confirmation (default)
```

---

#### `verbose`

Use this boolean flag to toggle verbose mode on or off. Verbose mode will print information at every step to help diagnose problems.

```python
interpreter.verbose = True  # Turns on verbose mode
interpreter.verbose = False  # Turns off verbose mode
```

---

#### `max_output`

This property sets the maximum number of tokens for the output response.

```python
interpreter.max_output = 2000
```

---

#### `conversation_history`

A boolean flag to indicate if the conversation history should be stored or not.

```python
interpreter.conversation_history = True  # To store history
interpreter.conversation_history = False  # To not store history
```

---

#### `conversation_filename`

This property sets the filename where the conversation history will be stored.

```python
interpreter.conversation_filename = "my_conversation.json"
```

---

#### `conversation_history_path`

You can set the path where the conversation history will be stored.

```python
import os
interpreter.conversation_history_path = os.path.join("my_folder", "conversations")
```

---

#### `model`

Specifies the language model to be used.

```python
interpreter.llm.model = "gpt-3.5-turbo"
```

---

#### `temperature`

Sets the randomness level of the model's output.

```python
interpreter.llm.temperature = 0.7
```

---

#### `system_message`

This stores the model's system message as a string. Explore or modify it:

```python
interpreter.system_message += "\nRun all shell commands with -y."
```

---

#### `context_window`

This manually sets the context window size in tokens.

We try to guess the right context window size for you model, but you can override it with this parameter.

```python
interpreter.llm.context_window = 16000
```

---

#### `max_tokens`

Sets the maximum number of tokens the model can generate in a single response.

```python
interpreter.llm.max_tokens = 100
```

---

#### `api_base`

If you are using a custom API, you can specify its base URL here.

```python
interpreter.llm.api_base = "https://api.example.com"
```

---

#### `api_key`

Set your API key for authentication.

```python
interpreter.llm.api_key = "your_api_key_here"
```

---

#### `max_budget`

This property sets the maximum budget limit for the session in USD.

```python
interpreter.max_budget = 0.01 # 1 cent
```