pinecone

Manage vector indexes, upserts, and hybrid queries with the Pinecone API.

1|Updated Mar 12, 2026

Search Tags:#serverless #semantic search #embeddings #rag #vector database #hybrid search #pinecone

Ready to run

npx skills add https://github.com/kaminocorp/hermes-alpha-hunter --skill pinecone-kaminocorp

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pinecone
Source: https://github.com/kaminocorp/hermes-alpha-hunter/tree/main/skills/mlops/vector-databases/pinecone
Command: npx skills add https://github.com/kaminocorp/hermes-alpha-hunter --skill pinecone-kaminocorp

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pinecone-client, and includes references (resource) components.

What problem does it solve? Building production RAG, semantic search, or recommendation systems requires a vector database that scales without infrastructure management, and this Skill provides the operational knowledge to use Pinecone's managed service correctly. ## Core Features & Use Cases - Index Management: Create serverless or pod-based indexes, upsert vectors in batches, and query with metadata filtering and namespaces. - Hybrid Search: Combine dense and sparse vectors with an alpha parameter for semantic plus keyword retrieval. - Framework Integration: Connect Pinecone to LangChain and LlamaIndex as a vector store for RAG pipelines. - Use Case: Build a multi-tenant RAG application where each customer's documents are isolated in namespaces, queried with metadata filters, and served with sub-100ms p95 latency. ## Quick Start Set up a Pinecone serverless index and show me how to upsert document embeddings and query them with metadata filters.

Frequently Asked Questions about pinecone

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I create a Pinecone index in Python?▼

Create a Pinecone index by initializing the Pinecone client with your API key and calling create_index with a name, dimension matching your embedding model, metric such as cosine, and a ServerlessSpec or PodSpec. Then connect with pc.Index(name) to start upserting vectors.

How to filter Pinecone query results by metadata?▼

Pinecone metadata filtering uses a filter parameter in the query call supporting operators like $eq, $gte, $lte, $in, $and, and $or. Attach metadata during upsert, then pass filters such as {"category": "tutorial"} to narrow results.

Pinecone vs Chroma vs FAISS for vector search?▼

Pinecone is a fully managed, auto-scaling SaaS suited for production workloads with low latency requirements. Chroma is self-hosted and open-source, while FAISS is an offline library for pure similarity search without a server component.

Does Pinecone support hybrid search with sparse vectors?▼

Yes, Pinecone supports hybrid search by upserting both dense values and sparse_values with token indices and weights. Query with both vector and sparse_vector parameters and tune the alpha value between 0 (sparse) and 1 (dense).

When should I use Pinecone namespaces?▼

Use Pinecone namespaces to isolate data by user or tenant in multi-tenant applications, or to separate environments like staging and production. Pass the namespace parameter during upsert and query to keep partitions independent.