17 KiB

Raw Blame History

Welcome to Swarms Docs Home

What is Swarms?

Swarms is the first and most reliable multi-agent production-grade framework designed to orchestrate intelligent AI agents at scale. Built for enterprise applications, Swarms enables you to create sophisticated multi-agent systems that can handle complex tasks through collaboration, parallel processing, and intelligent task distribution.

Key Capabilities

🏢 Production-Ready: Enterprise-grade infrastructure with high reliability, comprehensive logging, and robust error handling
🤖 Multi-Agent Orchestration: Support for hierarchical swarms, parallel processing, sequential workflows, and dynamic agent rearrangement
🔄 Flexible Integration: Multi-model support, custom agent creation, extensive tool library, and multiple memory systems
📈 Scalable Architecture: Concurrent processing, resource management, load balancing, and horizontal scaling capabilities
🛠️ Developer-Friendly: Simple API, extensive documentation, active community, and CLI tools for rapid development
🔐 Enterprise Security: Built-in error handling, rate limiting, monitoring integration, and audit logging

Why Choose Swarms?

Swarms stands out as the most reliable multi-agent framework because it was built from the ground up for production environments. Unlike other frameworks that focus on research or simple demos, Swarms provides the infrastructure, tooling, and best practices needed to deploy multi-agent systems in real-world applications.

Whether you're building financial analysis systems, healthcare diagnostics, manufacturing optimization, or any other complex multi-agent application, Swarms provides the foundation you need to succeed.

Get started learning swarms with the following examples and more.

Install 💻

$ pip3 install -U swarms

Using uv (Recommended)

uv is a fast Python package installer and resolver, written in Rust.

# Install uv
$ curl -LsSf https://astral.sh/uv/install.sh | sh

# Install swarms using uv
$ uv pip install swarms

Using poetry

# Install poetry if you haven't already
$ curl -sSL https://install.python-poetry.org | python3 -

# Add swarms to your project
$ poetry add swarms

From source

# Clone the repository
$ git clone https://github.com/kyegomez/swarms.git
$ cd swarms

# Install with pip
$ pip install -e .

Environment Configuration

Learn more about the environment configuration here

OPENAI_API_KEY=""
WORKSPACE_DIR="agent_workspace"
ANTHROPIC_API_KEY=""
GROQ_API_KEY=""

🤖 Your First Agent

An Agent is the fundamental building block of a swarm—an autonomous entity powered by an LLM + Tools + Memory. Learn more Here

from swarms import Agent

# Initialize a new agent
agent = Agent(
    model_name="gpt-4o-mini", # Specify the LLM
    max_loops=1,              # Set the number of interactions
    interactive=True,         # Enable interactive mode for real-time feedback
)

# Run the agent with a task
agent.run("What are the key benefits of using a multi-agent system?")

🤝 Your First Swarm: Multi-Agent Collaboration

A Swarm consists of multiple agents working together. This simple example creates a two-agent workflow for researching and writing a blog post. Learn More About SequentialWorkflow

from swarms import Agent, SequentialWorkflow

# Agent 1: The Researcher
researcher = Agent(
    agent_name="Researcher",
    system_prompt="Your job is to research the provided topic and provide a detailed summary.",
    model_name="gpt-4o-mini",
)

# Agent 2: The Writer
writer = Agent(
    agent_name="Writer",
    system_prompt="Your job is to take the research summary and write a beautiful, engaging blog post about it.",
    model_name="gpt-4o-mini",
)

# Create a sequential workflow where the researcher's output feeds into the writer's input
workflow = SequentialWorkflow(agents=[researcher, writer])

# Run the workflow on a task
final_post = workflow.run("The history and future of artificial intelligence")
print(final_post)

🏗️ Multi-Agent Architectures For Production Deployments

swarms provides a variety of powerful, pre-built multi-agent architectures enabling you to orchestrate agents in various ways. Choose the right structure for your specific problem to build efficient and reliable production systems.

Architecture	Description	Best For
SequentialWorkflow	Agents execute tasks in a linear chain; one agent's output is the next one's input.	Step-by-step processes like data transformation pipelines, report generation.
ConcurrentWorkflow	Agents run tasks simultaneously for maximum efficiency.	High-throughput tasks like batch processing, parallel data analysis.
AgentRearrange	Dynamically maps complex relationships (e.g., `a -> b, c`) between agents.	Flexible and adaptive workflows, task distribution, dynamic routing.
GraphWorkflow	Orchestrates agents as nodes in a Directed Acyclic Graph (DAG).	Complex projects with intricate dependencies, like software builds.
MixtureOfAgents (MoA)	Utilizes multiple expert agents in parallel and synthesizes their outputs.	Complex problem-solving, achieving state-of-the-art performance through collaboration.
GroupChat	Agents collaborate and make decisions through a conversational interface.	Real-time collaborative decision-making, negotiations, brainstorming.
ForestSwarm	Dynamically selects the most suitable agent or tree of agents for a given task.	Task routing, optimizing for expertise, complex decision-making trees.
SpreadSheetSwarm	Manages thousands of agents concurrently, tracking tasks and outputs in a structured format.	Massive-scale parallel operations, large-scale data generation and analysis.
SwarmRouter	Universal orchestrator that provides a single interface to run any type of swarm with dynamic selection.	Simplifying complex workflows, switching between swarm strategies, unified multi-agent management.

SequentialWorkflow

A SequentialWorkflow executes tasks in a strict order, forming a pipeline where each agent builds upon the work of the previous one. SequentialWorkflow is Ideal for processes that have clear, ordered steps. This ensures that tasks with dependencies are handled correctly.

from swarms import Agent, SequentialWorkflow

# Initialize agents for a 3-step process
# 1. Generate an idea
idea_generator = Agent(agent_name="IdeaGenerator", system_prompt="Generate a unique startup idea.", model_name="gpt-4o-mini")
# 2. Validate the idea
validator = Agent(agent_name="Validator", system_prompt="Take this startup idea and analyze its market viability.", model_name="gpt-4o-mini")
# 3. Create a pitch
pitch_creator = Agent(agent_name="PitchCreator", system_prompt="Write a 3-sentence elevator pitch for this validated startup idea.", model_name="gpt-4o-mini")

# Create the sequential workflow
workflow = SequentialWorkflow(agents=[idea_generator, validator, pitch_creator])

# Run the workflow
elevator_pitch = workflow.run()
print(elevator_pitch)

ConcurrentWorkflow (with `SpreadSheetSwarm`)

A concurrent workflow runs multiple agents simultaneously. SpreadSheetSwarm is a powerful implementation that can manage thousands of concurrent agents and log their outputs to a CSV file. Use this architecture for high-throughput tasks that can be performed in parallel, drastically reducing execution time.

from swarms import Agent, SpreadSheetSwarm

# Define a list of tasks (e.g., social media posts to generate)
platforms = ["Twitter", "LinkedIn", "Instagram"]

# Create an agent for each task
agents = [
    Agent(
        agent_name=f"{platform}-Marketer",
        system_prompt=f"Generate a real estate marketing post for {platform}.",
        model_name="gpt-4o-mini",
    )
    for platform in platforms
]

# Initialize the swarm to run these agents concurrently
swarm = SpreadSheetSwarm(
    agents=agents,
    autosave_on=True,
    save_file_path="marketing_posts.csv",
)

# Run the swarm with a single, shared task description
property_description = "A beautiful 3-bedroom house in sunny California."
swarm.run(task=f"Generate a post about: {property_description}")
# Check marketing_posts.csv for the results!

AgentRearrange

Inspired by einsum, AgentRearrange lets you define complex, non-linear relationships between agents using a simple string-based syntax. Learn more. This architecture is Perfect for orchestrating dynamic workflows where agents might work in parallel, sequence, or a combination of both.

from swarms import Agent, AgentRearrange

# Define agents
researcher = Agent(agent_name="researcher", model_name="gpt-4o-mini")
writer = Agent(agent_name="writer", model_name="gpt-4o-mini")
editor = Agent(agent_name="editor", model_name="gpt-4o-mini")

# Define a flow: researcher sends work to both writer and editor simultaneously
# This is a one-to-many relationship
flow = "researcher -> writer, editor"

# Create the rearrangement system
rearrange_system = AgentRearrange(
    agents=[researcher, writer, editor],
    flow=flow,
)

# Run the system
# The researcher will generate content, and then both the writer and editor
# will process that content in parallel.
outputs = rearrange_system.run("Analyze the impact of AI on modern cinema.")
print(outputs)

SwarmRouter: The Universal Swarm Orchestrator

The SwarmRouter simplifies building complex workflows by providing a single interface to run any type of swarm. Instead of importing and managing different swarm classes, you can dynamically select the one you need just by changing the swarm_type parameter. Read the full documentation

This makes your code cleaner and more flexible, allowing you to switch between different multi-agent strategies with ease. Here's a complete example that shows how to define agents and then use SwarmRouter to execute the same task using different collaborative strategies.

from swarms import Agent
from swarms.structs.swarm_router import SwarmRouter, SwarmType

# Define a few generic agents
writer = Agent(agent_name="Writer", system_prompt="You are a creative writer.", model_name="gpt-4o-mini")
editor = Agent(agent_name="Editor", system_prompt="You are an expert editor for stories.", model_name="gpt-4o-mini")
reviewer = Agent(agent_name="Reviewer", system_prompt="You are a final reviewer who gives a score.", model_name="gpt-4o-mini")

# The agents and task will be the same for all examples
agents = [writer, editor, reviewer]
task = "Write a short story about a robot who discovers music."

# --- Example 1: SequentialWorkflow ---
# Agents run one after another in a chain: Writer -> Editor -> Reviewer.
print("Running a Sequential Workflow...")
sequential_router = SwarmRouter(swarm_type=SwarmType.SequentialWorkflow, agents=agents)
sequential_output = sequential_router.run(task)
print(f"Final Sequential Output:\n{sequential_output}\n")

# --- Example 2: ConcurrentWorkflow ---
# All agents receive the same initial task and run at the same time.
print("Running a Concurrent Workflow...")
concurrent_router = SwarmRouter(swarm_type=SwarmType.ConcurrentWorkflow, agents=agents)
concurrent_outputs = concurrent_router.run(task)
# This returns a dictionary of each agent's output
for agent_name, output in concurrent_outputs.items():
    print(f"Output from {agent_name}:\n{output}\n")

# --- Example 3: MixtureOfAgents ---
# All agents run in parallel, and a special 'aggregator' agent synthesizes their outputs.
print("Running a Mixture of Agents Workflow...")
aggregator = Agent(
    agent_name="Aggregator",
    system_prompt="Combine the story, edits, and review into a final document.",
    model_name="gpt-4o-mini"
)
moa_router = SwarmRouter(
    swarm_type=SwarmType.MixtureOfAgents,
    agents=agents,
    aggregator_agent=aggregator, # MoA requires an aggregator
)
aggregated_output = moa_router.run(task)
print(f"Final Aggregated Output:\n{aggregated_output}\n")

The SwarmRouter is a powerful tool for simplifying multi-agent orchestration. It provides a consistent and flexible way to deploy different collaborative strategies, allowing you to build more sophisticated applications with less code.

MixtureOfAgents (MoA)

The MixtureOfAgents architecture processes tasks by feeding them to multiple "expert" agents in parallel. Their diverse outputs are then synthesized by an aggregator agent to produce a final, high-quality result. Learn more here

from swarms import Agent, MixtureOfAgents

# Define expert agents
financial_analyst = Agent(agent_name="FinancialAnalyst", system_prompt="Analyze financial data.", model_name="gpt-4o-mini")
market_analyst = Agent(agent_name="MarketAnalyst", system_prompt="Analyze market trends.", model_name="gpt-4o-mini")
risk_analyst = Agent(agent_name="RiskAnalyst", system_prompt="Analyze investment risks.", model_name="gpt-4o-mini")

# Define the aggregator agent
aggregator = Agent(
    agent_name="InvestmentAdvisor",
    system_prompt="Synthesize the financial, market, and risk analyses to provide a final investment recommendation.",
    model_name="gpt-4o-mini"
)

# Create the MoA swarm
moa_swarm = MixtureOfAgents(
    agents=[financial_analyst, market_analyst, risk_analyst],
    aggregator_agent=aggregator,
)

# Run the swarm
recommendation = moa_swarm.run("Should we invest in NVIDIA stock right now?")
print(recommendation)

GroupChat

GroupChat creates a conversational environment where multiple agents can interact, discuss, and collaboratively solve a problem. You can define the speaking order or let it be determined dynamically. This architecture is ideal for tasks that benefit from debate and multi-perspective reasoning, such as contract negotiation, brainstorming, or complex decision-making.

from swarms import Agent, GroupChat

# Define agents for a debate
tech_optimist = Agent(agent_name="TechOptimist", system_prompt="Argue for the benefits of AI in society.", model_name="gpt-4o-mini")
tech_critic = Agent(agent_name="TechCritic", system_prompt="Argue against the unchecked advancement of AI.", model_name="gpt-4o-mini")

# Create the group chat
chat = GroupChat(
    agents=[tech_optimist, tech_critic],
    max_loops=4, # Limit the number of turns in the conversation
)

# Run the chat with an initial topic
conversation_history = chat.run(
    "Let's discuss the societal impact of artificial intelligence."
)

# Print the full conversation
for message in conversation_history:
    print(f"[{message['agent_name']}]: {message['content']}")

17 KiB Raw Blame History