Our method
How we test AI tools
We pre-select AI tools that have proven useful and test each one hands-on in our own account. This page explains how we choose, test and rate tools, how we keep reviews current, and how the site makes money.
The short version
Six things you can check on any review in our current format. We move older reviews to it as we update them.
We test it ourselves
On the free plan, the full trial, a plan we pay for, or access a company gives us. The review names which.
Every screenshot is ours
They come from our own test, small flaws included. We never use a company's marketing images as evidence.
We say what we didn't test
Near the top of each review, a How we tested box sums up the test. Its Not tested line lists what we left out.
Ratings are independent
No company can pay for a rating. Affiliate links and partnerships never change one.
Every rating has a context
We rate how useful a tool is for the people it's built for. The Best for line says who they are.
Prices are dated
Checked on the company's own pricing page, in US dollars, with the month we checked.
How we choose tools
We don't try to list every AI tool on the market. We keep a shortlist of tools that have earned their place in real work, and add a newcomer when it shows clear promise. The aim is a shorter list that saves you time, not the biggest directory.
A useful tool earns its place on our lists whether or not it has an affiliate program.
How a review is made
Each review goes through the same six steps, all on the same day.
We pick a tool
From our shortlist, in a category we cover: an established favorite or a promising newcomer.
We get access
We sign up and test on the free plan or the full trial. Some tools we pay for ourselves, because we use them in our own work every day. Now and then a company gives us a paid plan, usually for a month, so we can test the premium features. That access comes with no conditions on the rating, and the review says when we had it.
We give it a real job
We use the tool for a task from our own work and report what happened, including what didn't go to plan. In ClickUp, we tracked six reviews from our editorial pipeline. Hedra made a 15-second video ad for this site. Anara compared two research papers, and we checked its headline figures against the papers.
We capture what we see
The screenshots in a review come from that session, in the order you meet the screens, small flaws included. When a test involves private data, we don't show those screens.
We check the facts
Prices and plan limits come from the company's pricing page and help docs, product changes from its release notes. What users think comes from places like Reddit, community forums, Trustpilot and app stores, and we say where. Then we cross-check the key facts a second time with AI research tools such as Perplexity.
We rate it and publish
We set the rating, finish the review and publish it. The Last updated date at the top is that day.
What a review shows you
A few parts of each review in our current format tell you what it rests on. Here they are from our ClickUp review.
ClickUp review
By the Max Productive AI editorial team. Last updated 15 October 2026.
How we tested ClickUp
- Method
- Hands-on use, October 2026
- Access
- Our own Free Forever workspace, with the built-in Brain² trial
- Scope
- Board, tasks, a Doc wiki, and Brain² reports, subtasks and an Artifact dashboard
- Not tested
- Super Agents, automations, Gantt and Dashboards, integrations and importers, mobile apps, paid plans
- The rating label
- Next to the rating: what it rests on, such as “Hands-on test, free plan” or “Hands-on test, paid plan”. It links to the test.
- The How we tested box
- The same four lines on every review: Method, Access, Scope and Not tested. When a company gave us access, a line above them says so. Anything under Not tested relies on the company's documentation, and the review says that too.
- What changed since our last review
- On updated reviews, a dated list of product and price changes since our previous test.
- About this review
- At the end of the page: what the review is based on, the sources we checked and when.
How we rate
Each tool gets one overall rating out of 5. It's our judgment of the tool's practical benefit after the test: how well it does its job for the people it's made for.
Read a rating together with its Best for line. A tool can be excellent for one person and a poor fit for another. An AI app builder that lets a beginner build a working app without writing code can feel limiting to an experienced developer, and both views are fair. No tool is perfect, and every tool we list has people and jobs it suits well.
There is no formula and there are no sub-scores. The verdict, the pros and cons and the Best for line explain the number in words.
- 4.5 to 5.0
- Excellent
- A first choice for its audience.
- 4.0 to 4.4
- Very good
- Strong, with trade-offs worth knowing before you buy.
- 3.5 to 3.9
- Good
- A good fit for specific needs. Check the limits first.
- 3.0 to 3.4
- Fair
- Useful for a narrow need. Compare before you choose.
Why most of our ratings are high
Because we pre-select. Tools that don't hold up rarely reach a review, so most of the ratings you see are in the top two bands.
What never changes a rating
Our ratings are independent of any partnership with a brand or software company.
- A payment or a sponsorship
- An affiliate link or commission
- Free access a company gives us
- A request for a better rating
How we make money
Max Productive AI is free to read. Today, three things pay for it.
- Affiliate links
- Many reviews and articles link to the tools they cover. If you sign up or buy through one of those links, we may earn a commission at no extra cost to you. Reviews and articles with affiliate links say so at the top.
- Sponsored articles
- Some companies pay us to publish an article that features their product. A sponsorship never buys or changes a rating.
- Ads
- Some pages show ads from Google.
How we use AI
AI helps us research, analyze data and draft our pages, working from our own test notes and screenshots. The tests, the screenshots, the ratings and the opinions are ours, and a person checks and publishes every page.
What you read is what we found using the tools ourselves. We've been testing AI tools for more than two years, across hundreds of tools and very different jobs, and we write as experienced users rather than AI researchers.
How we keep reviews current
AI tools change fast, sometimes within a few weeks. When a tool ships a major release or a major new feature, we aim to test it again and update the review. Each update re-checks the prices and adds a dated line to What changed since our last review.
Every review shows the date of its latest test. If a tool has changed since then, we haven't caught up with it yet, and the company's own site has the latest.
About prices. Prices are in US dollars, for buyers in the United States. Other regions can pay different prices, and the company's own pricing page has the current one.
If we get something wrong
We check every page, but we're human and AI tools move fast, so mistakes happen. If you spot one, such as a wrong price, a missing major feature or out-of-date information, tell us on the contact page and we'll fix it.
Companies whose tools we cover can use the contact page too, for example to flag a wrong price or a missing feature. You can also tell us when you simply disagree with what we wrote.
Who we are
Max Productive AI is an independent site that reviews AI tools for work, published by the Max Productive AI editorial team. We have tested 500+ AI tools across our reviews and comparisons.
We review to show where each tool helps and where it falls short, so you can pick the one that fits you, never to knock a brand down. The field changes in weeks, which is a big part of why we enjoy it.
Changes to this page
- October 2026
- First published.