Tools · 10 min read

5 AI tools every small business should try before spending a dime.

The AI tools market right now is loud, crowded, and full of products that are one good landing page wrapped around a slightly repackaged version of something you could get for free. Before you buy a subscription to anything, there are five tools you should spend time with first. All free. All genuinely useful. And one afternoon with them will teach you more about how you actually want to use AI than six months of reading review articles.

A note on how to read this list

This is not a ranking. These five tools are not competing with each other, they cover different ground, and the right way to use them is to know which one to reach for when. I will tell you what each one is genuinely good at, where it tends to fall short, and the one thing worth trying first so you can form your own opinion fast.

I will also tell you when and why a paid upgrade is actually worth it, because the free tiers are real tools, not crippled demos, but the paid tiers are meaningfully better in specific situations. You should know which situations before you hand anyone a credit card.

1. Claude (claude.ai)

What it is: A conversational AI assistant made by Anthropic. If you have heard of ChatGPT, Claude is in the same category. It is the one I use most in my own work, which I wrote about last month, so I will try to be fair rather than just enthusiastic.

What it is genuinely great at: Long, nuanced writing tasks. Claude handles context unusually well, meaning you can give it a lot of background: a long email thread, a detailed situation, a complicated set of requirements, and it will hold all of that in mind as it responds. It also tends to produce prose that does not sound like it was written by a robot, which turns out to matter quite a bit when you are editing for tone. If you need to write something that has to feel human: a sensitive message to an employee, a letter to a difficult client, a proposal that needs to be persuasive rather than just correct, Claude is usually my first stop.

Where it falls short: The free tier has usage limits that you will hit if you use it heavily. It also does not browse the web in its standard mode, so anything requiring current information needs to be verified elsewhere.

First thing to try: Paste in a real email you need to write: give it the context, the recipient, the goal, and anything you specifically want to avoid saying, and ask for a draft. Compare the output to what you would have written on your own. That gap is your baseline.

Is the paid version worth it? If you hit the free limit within a week and find yourself annoyed by it, yes. The Pro plan removes the usage cap, adds file uploads, and unlocks the more capable model. For anyone using it as a real daily work tool, the cost pays for itself quickly.

2. ChatGPT (chat.openai.com)

What it is: The tool that put AI on the map for most people. Made by OpenAI. Still one of the most capable and widely supported AI assistants available.

What it is genuinely great at: Breadth. ChatGPT handles an enormous range of tasks competently: writing, analysis, coding, math, research, brainstorming, summarizing. It is also deeply integrated into a growing ecosystem of other tools, which matters if you are thinking about automation down the road. The free version now includes web browsing, which means it can pull current information rather than being limited to its training data. That is a meaningful advantage for anything where recency matters.

Where it falls short: The free tier uses an older model. The difference between the free and paid version is more noticeable here than with some of the other tools on this list. If a task requires real nuance or complex reasoning, you may find the free version giving you something that is technically correct but a little flat.

First thing to try: Ask it to explain something in your industry that you find genuinely complicated: a regulation, a process, a concept you have always understood intuitively but never been able to articulate clearly. Ask it to explain it to you as if you already know the basics, then ask a follow-up question. Notice how well it handles the back and forth.

Is the paid version worth it? If you are going to use AI heavily and want the best available model, yes. ChatGPT Plus gives you access to GPT-4o, which is meaningfully more capable. It is also the version worth having if you want to explore the plugin ecosystem or build custom GPTs for specific tasks.

3. Google Gemini (gemini.google.com)

What it is: Google’s AI assistant, built into the same account you probably already use for Gmail, Drive, and Docs.

What it is genuinely great at: Integration with things you already use. If your business runs on Google Workspace: Gmail, Docs, Sheets, Drive, Gemini is the tool most likely to fit naturally into the work you already do rather than sitting alongside it. It can summarize your email threads, help draft replies, pull information from your Drive, and assist inside documents without you switching tabs. For people who live in Google’s ecosystem, this is a significant practical advantage.

Where it falls short: As a standalone conversational tool, it is a step behind Claude and ChatGPT on nuanced writing and complex reasoning tasks. Where it shines is the integration story, not the raw capability story.

First thing to try: If you use Gmail, open a long email thread you have been avoiding and ask Gemini to summarize what has been decided and what still needs a response. That single use case alone will tell you whether the Google integration is valuable enough to make it a regular part of your workflow.

Is the paid version worth it? The paid tier, Gemini Advanced, makes more sense if you are on Google Workspace Business or Enterprise plans and want the AI features baked into your existing tools. If you are using the free Google account, try the free Gemini tier first and see how much you reach for it.

4. Microsoft Copilot (copilot.microsoft.com)

What it is: Microsoft’s AI assistant, powered by the same underlying technology as ChatGPT. Free to use, and if your organization runs Microsoft 365, a version of it is probably already in your environment.

What it is genuinely great at: The same integration story as Gemini, but for the Microsoft world. If your business uses Outlook, Teams, Word, and Excel, which describes a large portion of small and mid-sized businesses, Copilot embedded in those tools is worth knowing about. The standalone web version is also a capable general assistant with built-in web search, which makes it useful for quick research tasks where current information matters.

Where it falls short: The free standalone version is good but not exceptional. The genuinely impressive version, the one that works inside Word, Excel, Teams, and Outlook requires a Microsoft 365 Copilot license, which is a meaningful cost. The integration capabilities are real, but so is the price tag.

First thing to try: Use the free web version at copilot.microsoft.com and ask it to research something you have been meaning to look into: a competitor, a regulation, an industry trend. It will pull current web results and synthesize them. Compare that to doing the same research yourself and notice how much time it saves.

Is the paid version worth it? If your business is already deep in Microsoft 365 and you have specific workflows in Word, Excel, or Teams that eat significant time, a pilot of the Copilot license for a few users is worth evaluating. It is not an impulse buy, but the productivity case is real for the right organization.

5. Perplexity (perplexity.ai)

What it is: A different kind of AI tool. Less of a writing assistant, more of a research assistant. Perplexity answers questions by searching the web in real time and synthesizing the results, and it cites every source, so you can see exactly where the information came from.

What it is genuinely great at: Answering questions where accuracy and currency matter and you want to be able to verify what you are being told. This is the tool I reach for when I need to get up to speed on something quickly and I cannot afford to get it wrong. The source citations are not a nice-to-have: they are what make this tool trustworthy in a way that the others are not when you are doing genuine research.

Where it falls short: It is a research tool, not a writing tool. Do not ask it to draft your email or write your policy. That is not what it is built for. It answers questions and surfaces sources. The other tools on this list handle the creation side.

First thing to try: Ask it a question you actually need to know the answer to: something in your industry, a regulatory question, a market question. Read the answer, then click through to two of the sources it cites. Notice whether the synthesis is accurate. That exercise will build more trust in the tool faster than any demo.

Is the paid version worth it? The free version is genuinely capable. The paid version adds more searches, better models, and a few power-user features. If you find yourself using it daily for research, the upgrade is modest and worth it. Most people can get real value from the free tier.

How these five tools fit together

Here is the practical picture of how these tools work as a set, rather than as competitors:

  • For writing, drafting, and thinking out loud – Claude or ChatGPT. Try both, see which one feels more natural to you. Most people develop a preference within a week.
  • For research and getting current, sourced answers – Perplexity. Every time. The citations matter.
  • For work that lives inside your existing tools – Gemini if you are in Google Workspace, Copilot if you are in Microsoft 365. These earn their place through integration, not through being the best standalone tool.

You do not need all five running simultaneously. You need to know which one belongs in which situation. That knowledge comes from trying them, not from reading about them.

The one thing I want you to take away from this

The biggest mistake I see people make with AI tools is treating the research phase as a substitute for the doing phase. You can read reviews for a month and still not know whether Claude or ChatGPT is right for the way you work, because that answer depends on what you are actually doing and how you think. The only way to know is to use them.

Set aside two hours this week. Make free accounts for the ones on this list you have not tried yet. Bring a real task from your actual job, not a test task, a real one. Run it through two or three of these tools and compare the outputs. That two-hour investment will give you a clearer picture of where AI fits into your work than anything else you could do.

The best AI tool is the one you actually use. All five of these are free. There is no reason to have a strong opinion about any of them before you have spent time with them. Go find out.

A word on what comes next

Once you have a sense of which tools feel useful, the natural next question is how to use them better: specifically, how to ask better questions and get more consistent results. That is what December’s post is about: prompt engineering without the jargon, for people who just want AI to do what they are asking it to do. It is more practical than it sounds, and a few simple habits make a real difference.


This is post three of a two-year series on AI for real people doing real work. Post one covers what AI actually is. Post two is a look at how I use these tools in my actual job. Questions, or a tool you think belongs on this list? Send a note, I read every one.

Not sure which tools are right for your business?

The AI Automation Bundle starts with a workflow audit that maps your actual work to the tools that fit it best — no guessing, no wasted subscriptions, just a clear picture of what belongs in your stack.

Request a proposal →