Swarms API Documentation - Build AI Agents & Multi-Agent Systems

Stop copy-pasting docs. If you’re vibe coding in Cursor, automating with Claude Code, or building an agent pipeline with Codex, you can give your agent instant, accurate access to the entire Swarms API either through a live MCP server it can query on-demand, or a single llms.txt file you drop into context. Both are free, require no API key, and take under a minute to set up.

Overview

Two ways to give your coding agent full access to the Swarms API documentation:

Method	Best For	URL
MCP Server	Agents with MCP support (Claude Code, Cursor, Windsurf, etc.)	`https://docs.swarms.ai/mcp`
llms.txt	Paste-and-go context for any LLM or vibe coding session	`https://docs.swarms.ai/llms.txt`

No API key required for either. No installation needed.

Option 1: MCP Server (Recommended for Agents)

The Swarms Documentation MCP Server is a hosted remote server that exposes the entire Swarms API documentation as a live searchable tool. Connect it once and your agent can look up API references, code examples, and architecture guides on-demand. MCP Server URL: https://docs.swarms.ai/mcp

Available Tool: `SearchSwarmsApiDocumentation`

Field	Value
Operation ID	`MintlifyDefaultSearch`
Transport	HTTP

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "A query to search the content with."
    }
  },
  "required": ["query"]
}

Example queries your agent will run:

"agent completion API parameters"
"how to run a sequential workflow"
"mixture of agents example Python"
"authentication and API key setup"
"batch processing multiple agents"

Setup by Agent

Claude Code

claude mcp add --transport http swarms-docs https://docs.swarms.ai/mcp

Or add manually to your MCP config:

{
  "mcpServers": {
    "swarms-docs": {
      "url": "https://docs.swarms.ai/mcp"
    }
  }
}

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "swarms-docs": {
      "url": "https://docs.swarms.ai/mcp"
    }
  }
}

Cursor

Add to Cursor Settings → MCP, or to your project-level .cursor/mcp.json:

{
  "mcpServers": {
    "swarms-docs": {
      "url": "https://docs.swarms.ai/mcp"
    }
  }
}

Windsurf

Open the MCP configuration panel and add:

{
  "mcpServers": {
    "swarms-docs": {
      "url": "https://docs.swarms.ai/mcp"
    }
  }
}

OpenAI Codex / Agents SDK

from agents import Agent, Runner
from agents.mcp import MCPServerHTTP

swarms_docs = MCPServerHTTP(url="https://docs.swarms.ai/mcp")

agent = Agent(
    name="swarms-developer",
    instructions="You are an expert Swarms API developer. Use the Swarms documentation tool to look up accurate API details before writing code.",
    mcp_servers=[swarms_docs],
)

VS Code (GitHub Copilot)

Add to your workspace .vscode/settings.json:

{
  "mcp": {
    "servers": {
      "swarms-docs": {
        "type": "http",
        "url": "https://docs.swarms.ai/mcp"
      }
    }
  }
}

Continue.dev

Add to ~/.continue/config.json:

{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "transport": {
          "type": "http",
          "url": "https://docs.swarms.ai/mcp"
        }
      }
    ]
  }
}

Any MCP-Compatible Agent

Point any agent or framework that supports remote MCP HTTP transport at:

https://docs.swarms.ai/mcp

No authentication. No API key. Just works.

Option 2: llms.txt (For Vibe Coders)

llms.txt is a plain-text file containing the full Swarms API documentation in a format optimized for LLMs. Drop it into your context window, paste it into your system prompt, or use it as a reference file in any AI-assisted coding workflow. URL: https://docs.swarms.ai/llms.txt

When to Use llms.txt

Use llms.txt when:

You’re in a one-shot vibe coding session and want to paste docs directly into the prompt
Your agent or IDE doesn’t support MCP yet
You want to seed a new conversation with full API knowledge before asking your agent to write Swarms code
You’re using a raw API call (OpenAI, Anthropic, Groq, etc.) and want docs in the system prompt

How to Use It

In a system prompt:

Fetch https://docs.swarms.ai/llms.txt and use it as your reference for all Swarms API code you write.

In Claude / ChatGPT / Gemini:

Open https://docs.swarms.ai/llms.txt in your browser
Copy the full content
Paste it at the top of your conversation or into the system prompt field
Start asking questions and generating Swarms code with full doc context

In Cursor / Windsurf (as a context file):

@https://docs.swarms.ai/llms.txt

Add this reference at the top of your chat message. The agent will fetch and use the full docs as context. Via curl (download locally):

curl -o swarms-llms.txt https://docs.swarms.ai/llms.txt

Then reference the local file in your IDE or agent config.

MCP Server Specification

{
  "server": {
    "name": "Swarms API Documentation",
    "version": "1.0.0",
    "transport": "http"
  },
  "capabilities": {
    "tools": {
      "SearchSwarmsApiDocumentation": {
        "name": "SearchSwarmsApiDocumentation",
        "description": "Search across the Swarms API Documentation knowledge base to find relevant information, code examples, API references, and guides.",
        "inputSchema": {
          "type": "object",
          "properties": {
            "query": {
              "type": "string",
              "description": "A query to search the content with."
            }
          },
          "required": ["query"]
        },
        "operationId": "MintlifyDefaultSearch"
      }
    }
  }
}

Resources

Resource	Link
MCP Server	`https://docs.swarms.ai/mcp`
llms.txt	`https://docs.swarms.ai/llms.txt`
Swarms API Docs	docs.swarms.ai
Get API Key	swarms.world/platform/api-keys
Model Context Protocol	modelcontextprotocol.io
Discord Support	discord.gg/EamjgSaEQf

Getting Started

Clients

Agent Completions

Multi-Agent

Batch

Capabilities

Resources

Swarms Cloud

Docs MCP and LLMs txt

Overview

Option 1: MCP Server (Recommended for Agents)

Available Tool: `SearchSwarmsApiDocumentation`

Setup by Agent

Claude Code

Claude Desktop

Cursor

Windsurf

OpenAI Codex / Agents SDK

VS Code (GitHub Copilot)

Continue.dev

Any MCP-Compatible Agent

Option 2: llms.txt (For Vibe Coders)

When to Use llms.txt

How to Use It

MCP Server Specification

Resources

​Overview

​Option 1: MCP Server (Recommended for Agents)

​Available Tool: SearchSwarmsApiDocumentation

​Setup by Agent

​Claude Code

​Claude Desktop

​Cursor

​Windsurf

​OpenAI Codex / Agents SDK

​VS Code (GitHub Copilot)

​Continue.dev

​Any MCP-Compatible Agent

​Option 2: llms.txt (For Vibe Coders)

​When to Use llms.txt

​How to Use It

​MCP Server Specification

​Resources

Overview

Option 1: MCP Server (Recommended for Agents)

Available Tool: `SearchSwarmsApiDocumentation`

Setup by Agent

Claude Code

Claude Desktop

Cursor

Windsurf

OpenAI Codex / Agents SDK

VS Code (GitHub Copilot)

Continue.dev

Any MCP-Compatible Agent

Option 2: llms.txt (For Vibe Coders)

When to Use llms.txt

How to Use It

MCP Server Specification

Resources