Integration · Developer tools
Connect Claude & ChatGPT to the OpenAI API
Give your AI access to your own OpenAI API account. It can generate text with an OpenAI model of your choice, create embeddings, run content through moderation and search the files in your vector stores. It can also look up stored responses, files and batch jobs, so you can check on work you already run through the API.
Free plan, no credit card. Takes about a minute.
Try asking
“Ask gpt-5.4-mini to write three subject lines for this newsletter.”
The AI sends the prompt to the OpenAI Responses API with that model and returns the generated text.
“Search our product docs vector store for the refund policy.”
It runs a search on the vector store and returns the most relevant passages from your files.
“Check these five customer comments for harmful content.”
It sends the texts to the moderation endpoint and reports which ones are flagged and why.
What your AI can do
Text generation
Call the Responses and Chat Completions APIs with the model, instructions and tools you choose, and continue earlier conversations.
Embeddings and moderation
Create embedding vectors for text and check text against OpenAI's moderation categories.
Files and vector stores
List and inspect uploaded files, create vector stores, attach files and search them for relevant passages.
“Which models can our API key use?”
It lists the models available to your OpenAI account.
“How is the batch job from yesterday doing?”
It lists your batch jobs and reads the status and request counts of the one you mean.
“Add the uploaded price list file to our support vector store.”
It attaches the already-uploaded file to the vector store so it gets indexed for search.
29 tools for OpenAI (ChatGPT)
These tools are switched on when you connect. You can switch any of them off, or require your approval before it runs.
add_conversation_itemsChanges dataAdd up to 20 items, such as user or assistant messages, to a conversation by `conversation_id`.add_file_to_vector_storeChanges dataAttach an already-uploaded file (`file_id`) to a vector store so it is chunked and indexed for search.cancel_batchMoves moneyCancel an in-progress batch job by `batch_id`. Requests already completed are still billed.create_chat_completionChanges dataGenerate a reply with the Chat Completions API from a `model` and a `messages` list of {role, content}. Usage is billed to your OpenAI account.create_conversationChanges dataCreate a conversation that can hold items across several responses; pass its ID as `conversation` when generating a response.create_embeddingChanges dataCreate embedding vectors for text in `input` (a string or list of strings) with an embedding `model` such as text-embedding-3-small. Usage is billed to your OpenAI account.create_vector_storeChanges dataCreate a vector store with a `name` and optional already-uploaded `file_ids` to index for file search. Storage is billed to your OpenAI account.delete_conversationDeletesDelete a conversation by `conversation_id`. Per the API, the items in it are not deleted. This cannot be undone.delete_fileDeletesPermanently delete an uploaded file by `file_id`. This cannot be undone.delete_responseDeletesPermanently delete a stored response by `response_id`. This cannot be undone.generate_responseChanges dataGenerate text with an OpenAI model via the Responses API: pass `model` and `input` (text or message items), optionally `instructions`, `tools` like web_search, or `previous_response_id` to continue a thread. Usage is billed to your OpenAI account.get_batchGet one batch job by `batch_id`, including status and output file IDs.get_chat_completionGet a stored chat completion by `completion_id`.get_conversationGet a conversation by `conversation_id`, including its metadata.get_fileGet metadata for an uploaded file by `file_id`: name, size, purpose and status.get_modelGet basic details for one model by `model` ID.get_responseGet a stored response by `response_id`, including its output text, status and token usage.get_vector_storeGet one vector store by `vector_store_id`.list_batchesList your Batch API jobs with their status and request counts.list_chat_completionsList stored chat completions (only those created with store=true), optionally filtered by model or metadata.list_conversation_itemsList the items (messages, tool calls, outputs) in a conversation by `conversation_id`.list_filesList files uploaded to your OpenAI account, optionally filtered by `purpose` (assistants, batch, fine-tune, vision, user_data).list_modelsList the models available to your OpenAI account, with their IDs and owners.list_response_input_itemsList the input items that were sent to create a stored response, by `response_id`.list_vector_store_filesList the files attached to a vector store by `vector_store_id`, with their indexing status.list_vector_storesList your vector stores (knowledge bases used by file search), with file counts and status.moderate_contentChanges dataCheck text (or text and image inputs) against OpenAI's moderation categories and get flagged categories and scores.remove_file_from_vector_storeDeletesRemove a file from a vector store by `vector_store_id` and `file_id`. The file itself is not deleted.search_vector_storeChanges dataSearch the indexed files in a vector store by `vector_store_id` with a natural-language `query` and get the most relevant chunks.
Set up in three steps
- 1
Pick the app
Create a free PipMCP account and choose OpenAI (ChatGPT) from the app list.
- 2
Paste your key
Paste an OpenAI API key. Sign in at platform.openai.com, open Settings > API keys and click Create new secret key, then copy the key, which is shown only once. API usage is billed to your OpenAI API account, separately from any ChatGPT subscription.
- 3
Add the link to your AI
You get a personal MCP link. Add it to Claude, ChatGPT or Cursor:
- Click your name, then Settings › Connectors › Add custom connector.
- Paste your link as the Remote MCP server URL.
- Switch it on from the + menu in a chat.
Questions
What can the AI do in the OpenAI API?
It can generate responses and chat completions, create embeddings, run moderation checks, list models, and work with stored responses, conversations, files, vector stores and batch jobs. It can create vector stores and conversations, attach files, and delete files, responses and conversations.
Does the AI see my OpenAI API key?
No. Your OpenAI API key is encrypted at rest and never shown to the AI. After you save it, it is not shown again, not even to you. The AI only sees the results of the tools it calls.
Can I control what the AI is allowed to do?
Yes. You choose which tools are switched on. Generating text and embeddings uses paid API credit, and deleting files or responses is permanent, so these tools can require your approval before they run. Every tool call is logged.
Does it work with ChatGPT?
Yes. In ChatGPT go to Settings › Apps & Connectors › Advanced and turn on Developer mode, then add your PipMCP link. Developer mode needs a paid ChatGPT plan: Plus, Pro, Business or Enterprise. The same link also works in Claude (Settings › Connectors › Add custom connector), Cursor and other MCP clients.
What does it cost?
PipMCP has a free plan with no credit card required. Paid plans bill per completed task. OpenAI bills your API usage to your own OpenAI account.
Is this the same as my ChatGPT subscription?
No. This connects the OpenAI API with an API key from platform.openai.com. API usage is billed separately from a ChatGPT Plus, Pro, Business or Enterprise subscription, and it does not give access to your ChatGPT chat history.
Related integrations
Let your AI work in OpenAI (ChatGPT) today.
Start free. Your key stays encrypted, and you decide what the AI may do.
Connect OpenAI (ChatGPT) freePipMCP is not affiliated with OpenAI (ChatGPT). Product names are trademarks of their owners.






