# FAQs

## Which AI model should I use?

It depends on your use cases but generally use less capable models for simple tasks, and more powerful models for tasks that requires deeper reasoning

For example, simple tasks such as grammar fix... use `gpt-3.5-turbo` or local models like `mistral` or `llama2`

For more challenging tasks such as code refactoring, or finding bugs in code... use `gpt-4`, or `gpt-4-turbo`