A powerful VS Code extension for managing Weaviate vector databases with an intuitive GraphQL interface. Supports self-hosted and cloud Weaviate instances.
- Unified Experience: Manage, query, and monitor Weaviate directly in VS Code—no browser switching.
- Works Everywhere: Connect to local, on-prem, or cloud Weaviate instances.
- Visual Data Exploration: Interactive data browser with filtering, vector search, and export. Browse schema, run queries, and view results in rich tables and JSON.
- Intelligent GraphQL Editor: Auto-complete (wip), schema-aware templates, and error highlighting.
- Secure & Productive: Secure credential storage, hot reload, and type-safe development.
| VS Code | Cursor | Windsurf | Manual Install |
|---|---|---|---|
| Marketplace | Open VSX | Open VSX | Download VSIX |
- Multiple Connections: Manage several Weaviate instances at once
- Generative Search: Ask natural-language questions across one or more collections—configure top-k results per collection, view source-attributed context objects, and get combined LLM answers
- Data Explorer: Interactive visual browser with advanced filtering, 4 vector search modes (text, object, vector, hybrid), and JSON/CSV export
- RBAC & Security: Manage users, roles, and groups with native RBAC support and API key rotation
- Read-Only Mode: Connection-level guards to prevent accidental modifications to production data
- Backup & Restore: Create, monitor, and restore backups across multiple backends (filesystem, S3, GCS, Azure)
- Cluster Management: Comprehensive cluster information panel with health monitoring
- Schema Explorer: Visualize and browse collections, properties, and nested object structures
- Advanced Query Editor: Monaco-powered GraphQL editor with intelligent, schema-aware templates
- Live Results: Table and JSON views for query results
- Secure Storage: Credentials stored with VS Code's secret storage
- Modern UI: Responsive, dark-theme optimized interface
Try our quick local sandbox with Docker! Perfect for learning and development:
cd sandbox
docker-compose up -d
python3 populate.pyThis spins up a fully-configured Weaviate instance with sample jeopardy questions, vector embeddings, and backup support enabled. Learn more →
- Install Weaviate Studio from the VS Code Marketplace or Open VSX.
- Open the Weaviate Studio panel from the VS Code sidebar.
- Add a connection to your Weaviate instance (local, on-prem, or cloud).
- Explore your schema, run queries, and manage your data—all within VS Code!
Full Feature List
- Connect to multiple Weaviate instances simultaneously
- Secure credential storage with VS Code's built-in secret storage
- Connection health monitoring and automatic reconnection
- Cloud and Custom endpoints with API key support for Cloud
- Advanced options: per-operation timeouts and skip-initialization checks
- API keys are not displayed when editing existing connections
- Optional connect‑on‑expand prompt for disconnected connections
- Interactive table browser with sortable columns and pagination
- Visual filter builder with 10+ operators and AND/OR logic
- Four vector search modes: Text, Object, Vector, and Hybrid (BM25 + semantic)
- Alpha slider for balancing keyword vs semantic search
- Export to JSON/CSV with flexible scopes (current page, filtered results, or entire collection)
- Keyboard shortcuts (Ctrl+F, Ctrl+K, Ctrl+E) for power users
- User preferences persistence per collection
- Virtual scrolling for large datasets (1000+ objects)
- See Data Explorer README for full details
- Manage roles, users, and groups with native Weaviate RBAC support
- Create and edit complex role permissions with intuitive UI
- User management with API key rotation and OIDC password support
- Status tracking for active and inactive users
- Integrated into the connection management workflow
- Enable connection-level read-only guards to protect production data
- Visual indicators in the tree view for read-only connections
- Prevents destructive and creative operations (Add/Delete/Modify) across all studio tools
- Toggle state persisted per connection
- Chat-style interface for Retrieval-Augmented Generation queries
- Multi-collection support: Select one or more collections as pill badges; auto-add on selection from dropdown
- Query controls:
- Top results per collection: Configure how many objects to retrieve from each collection (3, 5, 10, 20)
- Query timeout: Adjustable timeout (30s, 60s, 2min, 5min) for long-running queries with local vectorizers
- Answer tools:
- Copy button: Copy full answer to clipboard
- Markdown toggle: Switch between formatted view and raw markdown source
- Query metadata: Shows which collections were queried and execution time (e.g., "From: Books, PodcastSearch • Completed in 2m 34s")
- Summary label ("3 collections selected") when 3+ collections are active to reduce visual clutter
- Validation hint ("Select at least one collection to start") when no collection is selected; Ask button disabled until valid
- Only RAG-capable collections (those with a generative module configured) are shown
- Right-click a collection → "Generative Search" pre-selects that collection
- Opening Generative Search from another collection while the panel is already open adds it as a new pill
- Per-collection retrieval with non-blocking error handling
- Source attribution: context objects display their originating collection name as a badge label
- Retrieved Context section: Collapsible view grouped by collection, showing properties, UUID, and distance/certainty/score metrics
- Click telescope icon (🔭) on any context object to open it in the Data Explorer
- Relies on Weaviate's server-side generative config—no additional API keys needed in the extension
- Loading, error, and empty states matching the rest of the extension UX
- Clear chat button to reset conversation history
- Enter to send, Shift+Enter for newline
- Table view: Flattened, readable tables for nested JSON
- JSON view: Syntax-highlighted, collapsible JSON viewer
- Schema explorer: Interactive browsing of your Weaviate schema
- Real-time results: Live query execution with instant feedback
- Monaco Editor with full GraphQL syntax support
- Auto-completion and context-aware suggestions
- Intelligent, schema-aware query generation that adapts to your collection schema
- Dynamic query templates with auto-populated properties and accurate vector dimensions
- Real-time validation and error highlighting
- Enhanced query templates for all major Weaviate operations with comprehensive error handling
- Reference field support and type-safe generation
- Browse collections and their properties
- View detailed schema information, data types, and relationships
- Support for cross-references and nested object structures with recursive navigation
- Visual property type icons (text, number, boolean, date, object, geo coordinates, phone, blob)
- Visual representation of your data model with hierarchical nested property display
- Create collections via three paths: From scratch, Copy from existing, or Import from JSON schema
- Comprehensive Cluster Information Panel with real-time monitoring
- Auto-opens on connection by default (configurable)
- "Save and Connect" workflow for streamlined connection setup
- Cluster health and status monitoring
- Node information and statistics
- Create backups with real-time progress tracking
- Monitor backup status (in-progress, success, failed)
- Restore backups from any available backend
- Retry failed backups or cancel in-progress operations
- Multi-backend support: filesystem, S3, GCS, Azure
- Automatic detection of available backup modules
- Advanced configuration: include/exclude collections, custom paths
- Independent refresh controls for backups, collections, nodes, and metadata
- Connection-level: Cluster information panel, backups, modules, collections overview
- Collection-level: Properties (with nested object support), vectors (with count), inverted index, generative config, statistics, sharding, replication, multi‑tenancy
- Enhanced schema viewer with overview, properties, raw JSON, API equivalents, and creation scripts
- Hot reload for instant updates
- Full TypeScript support
- Modern, responsive UI
- Core: Basic Get, Vector Search, Semantic Search, Hybrid Search
- Advanced: Filter, Aggregation, Relationship, Sort, Explore
- Intelligent, schema-aware templates that dynamically adapt to your collection's actual schema
- Auto-populated properties based on data types (primitives, geo coordinates, references)
- Support for 15+ popular embedding models with automatic dimension detection
- Comprehensive error handling with graceful fallback system
- Visual decision tree and troubleshooting guide for 41 common mistakes
See the GraphQL Templates Guide for detailed usage, examples, and best practices: docs/GRAPHQL_TEMPLATES.md
- Delete All Collections (destructive) with double confirmation
Weaviate Studio collects anonymous usage telemetry (extension activation and error categories) to help improve the extension. We never collect prompts, queries, data, credentials, collection names, or personally identifiable information.
Telemetry requires dual consent — both VS Code telemetry and the weaviate.telemetry.enabled setting must be enabled. Disable either to opt out completely. All collected events are documented in telemetry.json.
We welcome contributions! Please see CONTRIBUTING.md for setup, development, and pull request guidelines.
Need a quick Weaviate instance for development or learning? Check out our local sandbox environment with pre-configured Docker setup and test data.
Weaviate Studio uses a modular architecture with external React components for enhanced UI:
- Add Collection UI: Powered by
weaviate-add-collection(Live Demo) - a standalone React component for creating, cloning, and importing collections - Extension Core: TypeScript-based VS Code extension
- Generative Search: Self-contained module (
src/rag-chat/) for generative search with its own panel, API wrapper, and React webview - Webviews: React-based UIs with Monaco editor integration
For details on updating external dependencies, see the Working with Dependencies section in CONTRIBUTING.md.
- Comprehensive unit and integration tests with Jest
- Strict TypeScript and linting for code quality
- See TESTING_GUIDE.md for details
- Issues: GitHub Issues
- Marketplace: VS Code Marketplace
- Open VSX: Open VSX Registry
- License: MIT (LICENSE)
Happy querying with Weaviate Studio! 🚀