OpenAI API Documentation

This is the complete API documentation for connecting to the OpenAI API using MindCloud's Universal API. If you are new to MindCloud, start with the Introduction to see how one API can sit in front of many different apps.

MindCloud is an integration company that offers one normalized API to access 3,100+ apps from a single place. It cleans up provider-specific quirks, keeps requests and responses consistent, and centralizes authentication so you do not need to manage separate flows for every app. Browse other apps.

Meet OpenAI: Use OpenAI to generate text, images, audio, embeddings, videos, vector-store workflows, evals, and fine-tuning jobs. Keep reading to get started.

Quickstart

Before you run your first OpenAI request, you need three things:

1. A MindCloud account — sign in or create an account at MindCloud.
2. A MindCloud API Key — create one in API Keys. Keep it on your server.
3. At least one OpenAI connection — create or choose one in Connections. Note its connectionId.

Every OpenAI action uses the same URL pattern:

POST https://connect.mindcloud.co/v2/universal/apps/openAi/actions/{actionSlug}/run

You need a MindCloud API Key and a OpenAI connection. Then call an action such as List Models:

curl --request POST \
  --url "https://connect.mindcloud.co/v2/universal/apps/openAi/actions/list-models/run" \
  --header "Authorization: Bearer $MINDCLOUD_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
  "connectionId": "$CONNECTION_ID",
  "arguments": {}
}'

Every response comes back in the same envelope, with your rows in a data array:

{
  "success": true,
  "data": [
    { "id": "1042", "name": "Ava" }
  ],
  "meta": {}
}

After your first request works, use Shape your requests to control arguments, pagination, filtering, fields, and errors.

Authenticate with a MindCloud API Key plus a connectionId for your OpenAI account. See OpenAI API Authentication for the full flow and how to get your connection credentials.

OpenAI actions

All 42 published actions for this OpenAI version. Each page documents the endpoint, arguments, and response controls.

• POSTAdd File To Vector Store
• PUTCancel Batch
• PUTCancel Response
• POSTCompact Response
• GETCount Response Input Tokens
• POSTCreate Batch
• POSTCreate Chat Completion
• POSTCreate Conversation
• POSTCreate Conversation Item
• POSTCreate Embedding
• POSTCreate Response
• POSTCreate Transcription
• POSTCreate Translation
• POSTCreate Vector Store
• DELETEDelete Conversation
• DELETEDelete Conversation Item
• DELETEDelete File
• DELETEDelete Response
• DELETEDelete Vector Store
• DELETEDelete Vector Store File
• POSTGenerate Image
• GETGet Conversation
• GETGet Conversation Item
• GETGet File
• GETGet File Content
• GETGet Response
• GETGet Vector Store
• GETGet Vector Store File
• GETGet Vector Store File Content
• GETList Conversation Items
• GETList Files
• GETList Models
• GETList Response Input Items
• GETList Vector Store Files
• GETList Vector Stores
• POSTModerate Input
• GETRetrieve Batch
• GETRetrieve Model
• GETSearch Vector Store
• PUTUpdate Conversation
• PUTUpdate Vector Store
• POSTUpload File

Related apps

More apps on the Universal API, behind the same REST shape. Or browse all apps.

Google SheetsGmailSlackGoogle CalendarGoogle DriveNotionGoogle FormsMailchimp