Skip to main content

Vector Databases

Vector databases are databases that enable embedding searches. They allow you to store many/long documents and retrieve them later using an LLM. Langdock currently supports the following providers:

Setting Up a Connection

When working with vector databases, you first need to set up a connection in the respective integration’s settings. Once the vector database is connected, it can be used in agents as an action or in chat with @. With each request, the model writes a request to the database and an embedding search is performed.

Knowledge Folders

A knowledge folder is a collection of several hundred files that can be used as knowledge. Knowledge folders are vector databases within the product, so smaller use cases don’t require setting up a separate vector database, and less technical users can work with many documents.

Capacity

Up to 1,000 files can be uploaded manually or via API.

Management

Users can fully manage the knowledge folder through the user interface (Integrations → Knowledge Folders), including:
  • Uploading files
  • Deleting files
  • Managing permissions
Automatic synchronization from integrations is not possible.
Sharing: You can share knowledge folders with individual users, groups, or the entire workspace.

API Access

Alternatively, users can programmatically upload, update, or delete files via the Knowledge Folder API. A link can be included to reference the original source in responses to users.

Comparison: Knowledge Folders vs. Custom Vector Database

Knowledge folders are suitable for less technical users and smaller use cases that need to be implemented quickly. For very large, valuable use cases with several thousand documents, a custom vector database is recommended, as settings can be adjusted to the use case, paragraph length, topics, etc.
FeatureKnowledge Folders (Langdock)Custom Vector Database
ManagementFully managed by Langdock, no infrastructure neededFull control over models, dimensions, retrieval parameters, updates, etc.
Default SettingsOptimized defaults (2,000-character chunks, 1536 dimensions, top-50 retrieval)No predefined settings
Data VolumeIdeal for up to 1,000 filesUnlimited data volumes and custom data structures possible
SetupQuick setup via Langdock UIOwn infrastructure, maintenance, and connection via Langdock Actions required
Access ControlGranular access control and UI management in LangdockAccess control must be implemented yourself