> ## Documentation Index
> Fetch the complete documentation index at: https://docs-v1.latitude.so/llms.txt
> Use this file to discover all available pages before exploring further.

# OpenAI

> Common questions about the OpenAI provider

## How do I set the OpenAI endpoint?

You can use either the classic OpenAI [Chat Completions](https://platform.openai.com/docs/api-reference/chat) or [Responses](https://platform.openai.com/docs/api-reference/responses) endpoints.

To select your preferred endpoint when creating the provider, follow these steps:

1. Go to the **Settings** section.
2. Click on **Providers**.
3. Select **OpenAI** from the list of providers.
4. Choose the endpoint you want to use.
   <img src="https://mintcdn.com/latitudellms/NYIRJRDqJuCM5TjZ/assets/new-openai-provider.png?fit=max&auto=format&n=NYIRJRDqJuCM5TjZ&q=85&s=b29c6fe2dbdefc76272484d6fc0b0938" alt="Create an OpenAI provider" width="650" height="802" data-path="assets/new-openai-provider.png" />

<Note>
  Once you select an endpoint, all prompts using this provider will use that
  same endpoint.
</Note>

### Responses API

#### OpenAI built-in tools

When using [Responses](https://platform.openai.com/docs/api-reference/responses), you can access OpenAI's [built-in tools](https://platform.openai.com/docs/guides/tools?api-mode=responses), such as web search:

```markdown {6-8} theme={null}
---
provider: my-openai-provider
model: gpt-4.1
temperature: 0.1
tools:
  - openai:
      - type: web_search_preview
        search_context_size: low
---
```

## How do I set the OpenAI thinking budget?

The latest OpenAI models have a configurable [thinking budget](https://platform.openai.com/docs/guides/reasoning?api-mode=responses) that allows you to control the amount of time the model spends reasoning to generate a response.

```markdown {4-5} theme={null}
---
provider: name-of-your-openai-provider-in-latitude
model: o4-mini
reasoningSummary: detailed
reasoningEffort: high
---
```

<Note>
  The **Chat Completions** API does not support reasoning summaries. Switch to
  the **Responses** API for a better experience.
</Note>