pinecone-db-workbench
v0.1.1
Published
Zero-config Pinecone DB Workbench — AI-native vector database GUI. Indexes browser, semantic search, metadata explorer, embeddings viewer & query console. Run with one `npx pinecone-db-workbench`
Maintainers
Keywords
Readme
Pinecone DB Workbench
The Pinecone Vector Database Workbench — Zero-Config CLI
Create indexes · Browse namespaces · Query vectors · Test RAG flows — with a single npx command.
Overview
Pinecone DB Workbench is a zero-config CLI tool for Pinecone vector databases. Launch a local web UI with a single npx command to connect to Pinecone Cloud, a local Pinecone-compatible gateway, or a manager endpoint to create indexes, inspect namespaces, run similarity queries, and debug raw API responses.
It is designed for AI engineers building semantic search, retrieval augmented generation, recommendation systems, and embedding-heavy applications.
Features
| | Capability | Description |
|---|---|---|
| 🔌 | Cloud & Local Connect | Connect to Pinecone Cloud or a local Pinecone-compatible manager endpoint |
| 🧭 | Index Explorer | List indexes with dimensions, metrics, type, host, and status |
| 🧱 | Index Management | Create and inspect indexes from the workbench |
| 🧬 | Vector Tools | Upsert, fetch, and inspect vector records by namespace |
| 🔍 | Query Playground | Run similarity queries with top K, vector input, values, and metadata options |
| 🧪 | RAG Testing | Quickly validate embeddings, namespaces, and retrieval payloads |
| 🧰 | Raw API Panel | Send custom requests to advanced Pinecone endpoints |
| 🚀 | Zero Config | No installation required — runs with npx |
Quick Start
Step 1 — Launch
npx pinecone-db-workbenchStep 2 — Choose an endpoint
| Mode | Example |
|---|---|
| Cloud | https://api.pinecone.io or your Pinecone API host |
| Local | http://localhost:5081 for a local manager or gateway |
Step 3 — Connect
Enter your endpoint and API key when required, then click Connect.
Usage
npx pinecone-db-workbench [options]Options
-u, --url <url>- Default Pinecone endpoint shown in the UI-p, --port <port>- Port to serve the UI on (default: first free from 7070)--host <host>- Host to bind to (default: 127.0.0.1)--api-key <token>- Prefill API key--no-open- Do not open the browser automatically-h, --help- Show help-v, --version- Show version
Examples
# Launch with default settings
npx pinecone-db-workbench
# Launch with specific endpoint
npx pinecone-db-workbench --url https://your-index.pinecone.io
# Launch with endpoint and API key
npx pinecone-db-workbench --url https://your-index.pinecone.io --api-key YOUR_KEY --port 9000Connection Modes
Cloud
Use Pinecone Cloud with an API endpoint and key. The workbench is built for live exploration of indexes, namespaces, vectors, and query responses.
https://api.pinecone.ioLocal
Use a local Pinecone-compatible manager, gateway, or development proxy.
http://localhost:5081This is useful for testing local workflows before connecting to production cloud resources.
Workbench Tabs
| Tab | Use it for | |---|---| | Dashboard | Connection status, selected index, and summary metrics | | Indexes | Browse, create, and inspect Pinecone indexes | | Vectors | Fetch, upsert, and inspect vectors in namespaces | | Query | Run similarity queries and review results | | Settings / Raw | Debug custom endpoint calls and manager APIs |
Requirements
- Node.js
^18.0.0or higher - Pinecone Cloud access or a local Pinecone-compatible manager endpoint
- API key for secured cloud or private deployments
Support & Feedback
| | | |---|---| | Website | Visit the public product page | | npm Package | npmjs.com/package/pinecone-db-workbench | | Privacy Policy | View policy |
