Skip to main content

​
What is Splox?

Splox is a multi-agent workflow builder where you create AI systems by connecting agents, tools, and logic visually. Chat with your agents or trigger them programmatically — combine 10,000+ tools and skills to build domain-specific AI agents.

Chat with Agents

Browse the agent marketplace and start chatting instantly

Build Workflows

Create and customize multi-agent workflows visually

​
Two Ways to Use Splox

For everyone — no building requiredGo to chat.splox.io or download the desktop app to browse the agent marketplace and start chatting with pre-built agents via text or voice.
  • Browse published agents in the marketplace
  • Chat via text or voice
  • No configuration needed
  • Available on web, Windows, macOS, and Linux

​
How it Works

1

Build Your Workflow

On app.splox.io, connect agents, tools, and logic nodes visually. Each workflow is a multi-agent system.
2

Add Tools & Skills

Choose from 10,000+ MCP tools and procedural skills to make your agents domain-specific.
3

Chat or Trigger

Chat with your agent on chat.splox.io or the desktop app, or trigger programmatically via the API.

​
Example: Agent with Tools

How it works:
  1. Start node initiates the workflow
  2. Agent processes your request and can call any connected tools
  3. Tool router dynamically selects the right tool based on the agent’s needs
  4. Chat interface lets you interact with the agent in real-time
For multi-agent workflows, connect multiple agents where each agent handles a specialized task - one agent can call another agent as part of the workflow.

​
Why Choose Splox?

Build systems where agents collaborate. One agent analyzes, another executes, a third validates - all connected visually.Example: Research agent → Writing agent → Editing agent pipeline.
Access MCP (Model Context Protocol) tools and procedural skills. Combine them to create highly specialized, domain-specific agents.Example: Financial analysis agent with market data tools + calculation skills.
Interact with workflows through a full chat interface. Side panel or full view - test and use your agents conversationally.Example: Chat with your research agent, ask follow-ups, iterate on results.
Build workflows visually, add custom logic when needed. No infrastructure to manage.Example: Connect nodes visually, write Python for complex transformations.

​
Node Types

Build workflows with these building blocks:
  • Start Node - Define workflow entry points and triggers
  • Agent Node - Autonomous AI agents with LLM + Tools loops
  • Tool Node - Execute custom functions or MCP tools
  • Switch Node - Conditional branching based on data or AI decisions
  • Merge Node - Combine data from multiple sources

​
Getting Started

Quickstart

Build your first workflow

Core Concepts

Understand the fundamentals

Agent Marketplace

Chat with pre-built agents

Community

Get help and share ideas

​
Next Steps

Ready to build?
  1. Quickstart Guide - Create your first workflow
  2. Workflows - Learn workflow fundamentals
  3. Nodes - Understand node types
  4. Tools - Connect your tools