Quickstart
TaskingAI is a developer-friendly cloud platform designed for building and running LLM agents for AI-native applications.
Understanding the Basics
TaskingAI structures modern AI applications into three fundamental modules:
- Model: Models serve as the brain of your AI application, they generate responses to inputs independently or with the aid of tools and retrievals.
- Retrieval: Retrievals store additional knowledge that models might not be equipped with.
- Tool: Tools are functionalities that empower AI assistants to interact with external resources and execute specific actions.
Currently, TaskingAI supports the following type of Agents:
- Assistant: Assistants are capable of autonomously utilizing various tools to achieve a specific goal. These assistants can be customized for a wide range of application needs, such as customer service, content generation, and more.
Hosts
TaskingAI offers two types of hosting options:
- Cloud Host: TaskingAI provides a cloud-hosted solution for users seeking quick deployment of their AI applications without the hassle of managing infrastructure. AI Applications hosted on the cloud are accessible via the TaskingAI API.
- Self-Host: For users who prefer deploying their AI applications on proprietary infrastructure, TaskingAI also offers a self-hosted option. This offers more control over the deployment environment.
SDKs and APIs
TaskingAI currently supports the following SDKs and APIs: