Skip to main content
Ragkit is a managed RAG platform that helps you build AI products quickly. Interact with Ragkit using the API and web-based dashboard.
The Ragkit API is in Beta. We do our best to fix bugs quickly, reach out through your assigned Slack channel, or [email protected].

​
Getting Access

Accessing the Ragkit platform involves creating an organization and then generating a secure API key.
1

Create an Organization

After signing in, make an organization for your team and invite your team members in the dashboard. Things like billing, rate limits, and data access are scoped to an organization.
Create org
2

Generate an API Key

An API key is used to securely access the Ragkit API as your organizaiton. Make sure to store it in a secure location. We never store your API key and it will not be shown again.
Create API key
3

Send API Requests

Now you’re ready to send requests! Keep reading to learn the basics of the Ragkit Platform, or head on over to the API documentation.
During the closed beta requests will fail until API access is granted. Request access at: https://ragkit.com/beta.

​
Platform Basics

1

Commands

The Ragkit platform is made up of commands than can be composed together to build powerful applications. Each command has a narrow purpose and adheres to the Unix philosophy.
2

Namespaces

All data in the platform is siloed according to a namespace. Data cannot cross a namespace boundary implicitly. Organizations may create many namespaces, and in multi-tenant applications it’s common to create a namespace for each tenant.Common commands:

NS_CREATE

Creates a new namespace.

NS_LIST

Lists all namespaces in an organization.
3

Agents

An agent is a way to group data and tool access. agents can be very specific (an expert with documents about a single domain), or broad (an agent that routes request to other agents in the namespace).Common commands:

AG_CREATE

Creates a new agent.

AG_REQUEST

Request output from an agent.
4

Documents

Documents contain information that you want to be processed and made available to a specific agent (Text, PDFs, Word Documents, Presentations, Website URLs, etc.).Common commands:

DOC_ADD

Makes a document available to an agent.

​
All Commands

To see all available commands head on over to the API documentation.