Three ways to use Orbit
From API-first developers to non-technical end users — Orbit meets you where you are.
Orbit Framework API
Define agents, tools, graphs, swarms, and workflows programmatically through a resource-oriented REST API. Full control for engineering teams embedding AI into their own products.
Orbit Platform
Create and manage agents through a visual interface. Write prompts, attach tools, compose orchestration patterns, and publish — all without writing code.
Orbit Workspace
AI-powered workspace where teams interact with published agents through a familiar conversational UI — streaming, file uploads, and collaboration included.
Agents, not nodes. Data, not code.
Agent as the Atomic Unit
In Orbit, the Agent is the single building block. An agent is an LLM with a system prompt, tools, and optional knowledge bases. Need more? Compose agents into Graphs, Swarms, or Workflows. One building block, three orchestration patterns.
Definitions in Data
Agents, tools, and orchestrations are stored as data — not hardcoded in application logic. Define resources via API or Platform and Orbit converts them into live execution at runtime.
Extend Without Modifying
Build custom integrations as external MCP servers or A2A agents. Deploy in your infrastructure, register with Orbit. Your extensions are orchestrated — zero core framework changes required.
Focus on agent logic. Orbit handles the rest.
How Jonas Fitness Built AI-Powered Knowledge Agents
Jonas Fitness used Orbit's API to build and deploy AI agents for both internal and external knowledge base chat. Teams get instant answers from internal documentation, while customers access a public-facing AI chat for product support.