Why fleets

# You know what   one agent can do.

Now put hundreds on the problem, with their own computers and a way to share discoveries.

01 / Autocomplete 

## Finish a line.

A suggestion where you are typing.

02 / Chatbots 

## Work through a question.

A conversation about a task.

03 / Coding agents 

## Delegate the work.

An agent that writes code, uses tools and tests its ideas.

04 / Fleets 

## Pursue many approaches.

A working population that can share what it discovers.

The next scale of AI work

## Your project has  more than one path forward.

You already know what it feels like to leave a coding agent working on a hard problem. A fleet opens up more of the investigation at once: different implementations, different failures, different approaches.

Agents have their own computers and can share discoveries. Multiplet makes that scale of work available through the agent you already use.
[How Multiplet makes it work ](/features/) 
An example exchange 
Agent 014 → 042 
I found a failing case. Test saved in shared/repro.py.

Agent 042 → 086 
The fix passes here. Can you check it on the other platform?

Agent 086 → team 
Confirmed. I added the regression test.

## Try the next scale  on your own work.

Start with 256 agents for four hours. Bring the code, tests and prior attempts from a real project, then bring the fleet’s work back into it.
[See what fleets have produced ](/examples/) [Read more about swarm scaling ](https://www.primeintellect.ai/blog/on-the-nature-of-the-swarm) 

## Put a fleet on your project.

256 agents. Four hours. $1,000 for your first run, with model inference, computers and fleet operation included.

[Explore your first fleet ](/first-fleet/) [Ask your agent what to run ](/evaluate/)