Agent or chatbot?
You talk to both in a chat. The difference is what happens after you press enter.
| Chatbot | AI agent | |
|---|---|---|
| You give it | A question | A goal |
| What it does | Writes an answer | Opens files, runs code, browses, changes things |
| Who does the work | You copy and paste | The agent does it |
| How long it takes | Seconds | Minutes to hours |
| What it needs | A chat window | A computer to work on |
Anthropic, the maker of Claude, defines agents as systems where LLMs dynamically direct their own processes and tool usage, maintaining control over how they accomplish tasks
. In plain words: the model chooses its own steps, and does not follow a script someone wrote in advance.
What agents do today
- Coding. The agent reads your project, changes files, runs the tests and reports back. Claude Code and Codex are agents of this kind.
- Research. It opens many pages, collects the facts and writes a summary with links.
- Routine jobs. It does the same chore every day, such as sorting a list or building a report.
Where an agent works
An agent needs a computer. On your laptop it works only while the laptop is awake and open. Close the lid and it stops. On a computer in the cloud it keeps going, and you read the result later from any browser, your phone included.
Do you need one?
Use an agent when
- The job has many steps. Read, change, test, fix, repeat.
- It takes longer than you want to watch. You give the task and come back later.
- You can check the result. A report, a page, a passing test.
Skip it when
- You need one answer. A chatbot is faster and costs nothing extra.
- Every step needs your judgment. An agent that must ask each time saves you nothing.
- A mistake cannot be undone. Do not let it touch money or the only copy of your data.
What can go wrong
- An agent can misread the goal and do the wrong thing with full confidence. Start with tasks you can check.
- It can get stuck, or loop on the same step, and need a nudge from you.
- Websites may block it. Many sites show captchas to traffic from data centers.
- On your own laptop its mistakes land on your own files. Give it a computer of its own, and a wrong step stays there.
Your agent gets a clean address of its own
Your computer runs in the Netherlands. Its traffic can leave through a static address of a regular internet provider in one of 8 countries. Only your account works from that address.
- From a data-center address
- Sites see a server range shared with thousands of strangers. The agent's browser hits more captchas and blocks.
- From your Clean exit IP
- Sites see a regular internet provider's address in the country you picked. Fewer captchas, and less reason for Claude or ChatGPT to flag your account: nobody else works from it.
The short way: an agent with a computer
On Uno you chat with your agent in Uno Work and it works on a Linux computer of its own. Close the laptop; the agent keeps going and the computer wakes on the first request.
Plus: $20 a month
A computer for your agent, from $5 a month
Every plan is a real Linux computer with root. Uno Work in the cloud, where you chat with your agents from any browser, comes with Plus.
Small
A Telegram bot, a VPN, a small backend
$5/ month
- 1 vCPU · 1 GB
- 15 GB disk
- 25 GB cloud storage
- 5 boosts a month
Plus
Agents that work with your laptop closed
$20/ month
- 4 vCPU · 4 GB
- 40 GB disk
- 100 GB cloud storage
- Uno Work in the cloud
Pro
A product in production, a second MacBook
$70/ month
- 16 vCPU · 16 GB
- 120 GB disk
- 300 GB cloud storage
- Speed of a MacBook Air M4
Max
A team and many agents at once
$200/ month
- 32 vCPU · 48 GB
- 300 GB disk
- 1,024 GB cloud storage
- 5 public IPv4
Add Uno AI to any plan: 40 h of AI work a month on Small, 100 h on Plus, unlimited on Max. Today: GLM-5.3 Flash and MiMo v2.6 Pro. Or sign in with your own Claude or ChatGPT.