# What is Pinecone? > Pinecone is a hosted vector database that stores embeddings so AI can search your documents by meaning. What it is, the free plan, and how n8n uses it. Source: https://learn.ynteractive.com/content/glossary/pinecone · Updated 2026-10-09 · Free from Stepthrough (https://learn.ynteractive.com) [n8n and automation](https://learn.ynteractive.com/content/glossary#topic-n8n-and-automation) · [AI glossary](https://learn.ynteractive.com/content/glossary) Quick answer Pinecone is a hosted vector database: a service that stores embeddings, the numeric fingerprints of your text, and returns the pieces closest in meaning to a question. Automation tools like n8n use it so an AI agent can search a handbook or a pile of documents before it answers. Last updated October 9, 2026 Updated October 9, 2026 · By [Robert Breen](https://learn.ynteractive.com/content/about) ## Why it matters for a small business A chatbot that answers from your own documents needs somewhere to keep them in a searchable form. Pinecone is one of the most common choices because there is nothing to install: you create an index in the web console, copy an API key, and tools such as n8n read and write to it. That makes it a practical first [vector database](https://learn.ynteractive.com/content/glossary/vector-database) for an office team building its first [RAG](https://learn.ynteractive.com/content/glossary/rag) chatbot. As of October 2026, the free **Starter** plan is enough to learn on: up to 5 indexes, with Serverless indexes on AWS in a handful of regions such as Virginia (us-east-1). Paid plans add more indexes, other clouds and higher limits. The thing to get right is the index setup: its **Dimension** must match the [embeddings](https://learn.ynteractive.com/content/glossary/embeddings) model you send, and you can't change it after the index is created. ## In a real lesson: Make Your Handbook Searchable for an AI Helpdesk using Pinecone + n8n In the [Pinecone + n8n load lesson](https://learn.ynteractive.com/content/pinecone-n8n-load-docs-vector-database) you run operations at Cedar Ridge Partners, a made-up consulting firm. In the Pinecone console you click **Create index**, name it **stepthrough-tests** and pick **text-embedding-3-small** under **Configuration**. Pinecone fills in **Metric: cosine**, but **Dimension** shows **512**, while n8n's **Embeddings OpenAI** node sends 1,536 numbers per chunk. You change it to **1536**, leave **Serverless**, **AWS** and **Virginia (us-east-1)**, and create the index. An n8n workflow then loads the employee handbook PDF into it through the **Pinecone Vector Store** node in **Insert Documents** mode, and the index browser shows 38 records, one per chunk. In the [chatbot lesson](https://learn.ynteractive.com/content/n8n-ai-agent-pinecone-rag-chatbot) the same node becomes a tool for an AI Agent, in **Retrieve Documents (As Tool for AI Agent)** mode with **Limit 4**, so every answer starts with a Pinecone search. [Try this lesson free](https://learn.ynteractive.com/modules/pinecone-n8n-load-docs) or [read the step-by-step guide](https://learn.ynteractive.com/content/pinecone-n8n-load-docs-vector-database). ## Common confusions ### Pinecone vs a vector database A [vector database](https://learn.ynteractive.com/content/glossary/vector-database) is the type of tool. Pinecone is one product of that type, hosted for you. Others include Qdrant, Weaviate and Supabase with pgvector, and n8n has nodes for several of them. ### Pinecone vs Pinecone Assistant The Pinecone console also offers an **Assistant** that takes uploaded files and answers questions itself. The lessons use the plain index, so the n8n agent, its rules and its citations stay under your control. ## Tips - Write down which embeddings model an index was built for, and use that same model to load and to search. - Use a separate index or namespace per document set, so a search never mixes the handbook with sales notes. - When you reload changed documents, clear the old records first or you'll store everything twice. ## Related terms [Vector database](https://learn.ynteractive.com/content/glossary/vector-database) · [Embeddings](https://learn.ynteractive.com/content/glossary/embeddings) · [Retrieval-augmented generation (RAG)](https://learn.ynteractive.com/content/glossary/rag) · [Chunking](https://learn.ynteractive.com/content/glossary/chunking) · [Semantic search](https://learn.ynteractive.com/content/glossary/semantic-search) · [Knowledge base](https://learn.ynteractive.com/content/glossary/knowledge-base) ## More n8n and automation terms [Pinned data](https://learn.ynteractive.com/content/glossary/pinned-data) · [Polling](https://learn.ynteractive.com/content/glossary/polling) · [Publishing a workflow](https://learn.ynteractive.com/content/glossary/publish-workflow) · [Rate limit](https://learn.ynteractive.com/content/glossary/rate-limit) ## Where you use it: free lessons - [Make Your Handbook Searchable for an AI Helpdesk using Pinecone + n8n](https://learn.ynteractive.com/content/pinecone-n8n-load-docs-vector-database) (n8n, 14 min) - [Answer Staff Questions from Your Own Docs using n8n + Pinecone](https://learn.ynteractive.com/content/n8n-ai-agent-pinecone-rag-chatbot) (n8n, 13 min) ## More free tutorials [n8n tutorials](https://learn.ynteractive.com/content/tool/n8n) ## Frequently asked questions **Is Pinecone free?** It has a free Starter plan, which as of October 2026 allows up to 5 indexes, Serverless on AWS, with monthly read, write and storage limits. Creating a sixth index asks you to upgrade to a paid plan. **What dimension should my Pinecone index use?** The number your embeddings model produces. OpenAI's text-embedding-3-small, n8n's default, produces 1,536, so choose 1536. If the numbers don't match, inserts fail. **Do I need to code to use Pinecone?** No. You create the index in the web console, and n8n's Pinecone Vector Store node handles inserting and searching with your API key saved as a credential. [All AI glossary terms, A to Z](https://learn.ynteractive.com/content/glossary) · [Free prompt templates](https://learn.ynteractive.com/content/prompts)