> ## Documentation Index
> Fetch the complete documentation index at: https://developers.reflection.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Base URLs, authentication, and conventions for the Reflection API

The Reflection API is an HTTPS API with JSON request and response bodies.

## Base URLs

| Endpoint | Base URL |
| - | - |
| Reflection client | `https://api.reflection.ai/v1` |
| OpenAI-compatible | `https://api.reflection.ai/openai/v1` |

The endpoints in this reference are served from the OpenAI-compatible base URL. See [OpenAI compatibility](/openai-compatibility) for how they relate to the OpenAI API.

## Authentication

Send your [API key](/authentication) in the `Authorization` header of every request:

```http theme={null}
Authorization: Bearer <your API key>
```

## Requests and responses

* Send request bodies as JSON with `Content-Type: application/json`. A body sent without that header returns `415`.
* Responses are JSON, except streamed chat completions, which use `text/event-stream`. See [Streaming](/streaming).
* Timestamps are Unix times in seconds.
* Errors return a JSON `error` object. See [Errors](/errors).
* Responses include an `x-request-id` header that identifies the request. See [Request IDs](/errors#request-ids).
* Responses can include `x-ratelimit-` headers that report your organization's remaining capacity. See [Rate limits](/rate-limits#rate-limit-headers).

## Endpoints

| Method | Path | Description |
| - | - | - |
| `POST` | `/openai/v1/chat/completions` | [Create a chat completion](/api-reference/chat/create-a-chat-completion) |
| `GET` | `/openai/v1/models` | [List models](/api-reference/models/list-models) |
| `GET` | `/openai/v1/models/{model}` | [Retrieve a model](/api-reference/models/retrieve-a-model) |