kenken999's picture
fda
0f43f8a
raw
history blame
1.88 kB
---
title: OpenAI
---
To use Open Interpreter with a model from OpenAI, simply run:
<CodeGroup>
```bash Terminal
interpreter
```
```python Python
from interpreter import interpreter
interpreter.chat()
```
</CodeGroup>
This will default to `gpt-4`, which is the most capable publicly available model for code interpretation (Open Interpreter was designed to be used with `gpt-4`).
<Info>
Trouble accessing `gpt-4`? Read our [gpt-4 setup
article](/language-model-setup/hosted-models/gpt-4-setup).
</Info>
To run a specific model from OpenAI, set the `model` flag:
<CodeGroup>
```bash Terminal
interpreter --model gpt-3.5-turbo
```
```python Python
from interpreter import interpreter
interpreter.llm.model = "gpt-3.5-turbo"
interpreter.chat()
```
</CodeGroup>
# Supported Models
We support any model on [OpenAI's models page:](https://platform.openai.com/docs/models/)
<CodeGroup>
```bash Terminal
interpreter --model gpt-4
interpreter --model gpt-4-32k
interpreter --model gpt-3.5-turbo
interpreter --model gpt-3.5-turbo-16k
```
```python Python
interpreter.llm.model = "gpt-4"
interpreter.llm.model = "gpt-4-32k"
interpreter.llm.model = "gpt-3.5-turbo"
interpreter.llm.model = "gpt-3.5-turbo-16k"
```
</CodeGroup>
# Required Environment Variables
Set the following environment variables [(click here to learn how)](https://chat.openai.com/share/1062cdd8-62a1-4aa8-8ec9-eca45645971a) to use these models.
| Environment Variable | Description | Where to Find |
| -------------------- | ---------------------------------------------------- | ------------------------------------------------------------------- |
| `OPENAI_API_KEY` | The API key for authenticating to OpenAI's services. | [OpenAI Account Page](https://platform.openai.com/account/api-keys) |