24 AI Tools Ranked from Essential to Forgettable for 2025
An honest review of which AI tools are actually worth your time
Dear subscribers,
Today, I want to share my tier list of the 24 most popular AI tools in the market.
If you want to try some new AI tools this holiday break, then this post is for you.
There are some controversial picks in this list, so I’ll also share my thoughts on where each tool excels and where it falls short. Let’s start at the bottom and work our way up.
This post is brought to you by…Shopify
Today, Shopify is launching 150+ updates to improve the performance, reliability, and integration of its existing products. I love how retro the Winter '25 Edition site is:
When you first land on it, you see “boring” static images and text.
But when you click the non-boring toggle, you’ll see more than 100 videos produced by AI tools that highlight nearly every update that Shopify is launching.
It’s a super fun and delightful website to play with. Check it out below:
D tier: Not useful nor fun
Let’s start with two controversial picks:
Apple Intelligence
Earlier this year, I wrote that Apple can win the personal AI race, assuming they could execute. Unfortunately, the reviews have been almost universally negative:
Even their headline feature - message summaries - produces unintentionally funny results that miss the point entirely:
Gemini for Google Workspace
I use Gmail, Google Calendar, and Docs daily, but I never use their AI features. I don’t want to use AI to:
Draft an email from scratch that’s not in my style.
Summarize a Google Doc in an unreadable side panel.
Generate AI backgrounds in Google Meet (who asked for this?)
It feels like the PMs who built these features don’t actually know what AI is good at.
Meanwhile, obvious AI use cases like summarizing comment threads in Docs are nowhere to be found. The wasted potential here is frustrating.
C-tier: Promising but not best-in-class
These tools show promise but aren’t the best in their category:
Figma AI isn’t ambitious enough. It pains me to write this as I love Figma, but their AI features lack ambition. There’s too much focus on basic tasks like renaming layers instead of the obvious use case of turning designs into prototypes.
Poe tries to do everything just OK. Getting access to the best AI models for $20 sounds great on paper. But Poe's clunky UX and lack of key features (no ChatGPT Voice or Claude Artifacts) make it feel like a compromise.
Gemini is the 3rd best LLM model. In my testing, Gemini consistently underperforms ChatGPT and Claude. Despite its supposedly large context window, responses are often too brief and frequently miss the point. Being the third-best LLM isn't great when the top two are just a click away.
Grammarly keeps suggesting the wrong things. I still use Grammarly, but it often makes suggestions that completely change a sentence's meaning. Grammarly could easily be in the A-tier if it worked more consistently.
B tier: Best-in-class but doesn’t fit my workflows
These tools excel at what they do, but I haven't found them essential for daily work:
Suno is best for creating songs. My kids and I use it for fun (it even works for other languages), but I haven't found a practical use case beyond entertainment.
HeyGen is best for talking head videos. AI creators like Rowan Cheung are using it to grow their social channels, but the videos still have an uncanny valley effect.
Websim is best for creating fun websites. I love creating websites from text ideas with my kids. It reminds me of the early internet’s creativity.
Grok is best for simple images. The quality might not be as good as Midjourney, but Grok is 100% free to use. I love using it with (you guessed it) my kids.
Playground AI is best for logo and social media design. I just wish it was better at creating designs with consistent branding (they’re working on it).
Napkin AI is best for infographics. The infographics look a bit too text-heavy to me, but it’s still an incredible tool.
Granola is best for meeting notes. This could be most employees' #1 time saver, so I hope Granola can get adopted by more companies soon.
Zapier AI is best for AI automation. Using AI to get work done without you being there is amazing. The UX is still clunky, so check out my guide for info.
Elevenlabs is best for text-to-speech and voice cloning. They also have a mobile app that lets you listen to text, articles, or PDFs in your favorite voices.
Runway is best for creating videos. You can generate great short clips in any style from text prompts. Reviews online say it’s better than OpenAI’s new Sora model.
NotebookLM is best for text to podcast. It can generate fun 10-minute podcast episodes. Although in practice, I don’t use it too much beyond entertainment.
A tier: Best in class and part of my workflow
These tools are not only best-in-class but also a key part of my workflow:
Superwhisper
The best way to improve AI prompts is to provide context, and voice is the fastest way to do that. I use this simple app extensively to write first drafts and capture ideas. Better Dictation is another solid option.
RiversideFM
RiversideFM is my go-to tool for video podcasts. Their AI features like auto-removal of silences and magic audio make post-production easy.
v0
This is what Figma's AI should have been. Drop in a mockup, and v0 creates a nearly pixel-perfect prototype. I use this regularly for rapid prototyping.
Cursor
Cursor is my preferred tool for coding as a not very technical person. At this point, you can create any simple tool using:
Claude/ChatGPT to write the requirements.
That being said, I still think it’s worth learning CS fundamentals first (I like Harvard’s CS50 videos) if you want to build anything beyond simple tools:
Check out my Cursor guide for a step-by-step walkthrough to build your first tool.
S tier: Tools that I use everyday
The S-tier includes the three usual suspects — Perplexity, ChatGPT, and Claude. I use all three tools daily, but let me explain the nuances of what they’re good at and not.
Perplexity
Perplexity has largely replaced Google Search for me. Whether I'm planning travel, researching topics, or looking up specific local information, it delivers accurate results with verifiable sources. For example, the other day, it gave me the exact answer to: “What is the registration date for spring activities for 6-year-olds in (my local neighborhood)?” Check out my Perplexity guide for advanced tips.
ChatGPT
The OG AI tool is still best-in-class at:
Answering quick questions. Claude runs out of tokens way too fast.
Voice interactions. I love talking to ChatGPT Voice on my daily commute.
Basic data analysis. I find ChatGPT more reliable at this than Claude.
They also offer the O1 model for solving complex problems and Sora for video generation (although the latter has mixed reviews so far). Paying $20 / month for all of the above is a no brainer.
Claude
Claude excels at writing and editing, making it my most-used AI tool. I’ve happily paid $20 / month for the past year.
However, Claude is experiencing growing pains with token limits and overly concise responses. It also lacks basic image and voice capabilities.
Overall, I use all 3 tools daily to get work done and follow my curiosity.
Wrap up
To wrap up, here’s the tier list again with links to each tool:
S: Perplexity, ChatGPT, Claude
A: Superwhisper, RiversideFM, v0, Cursor
B: Suno, HeyGen, Websim, Grok, Playground AI, Napkin AI, Granola, Zapier AI, Elevenlabs, Runway, NotebookLM
Let me know in the comments which tools you use most often and if you agree with these rankings!
I feel this list. I am disappointed in Apple, Google, Figma, and last mile solutions from code builders-yet looking forward to improvements. I am frustrated because I want to be part of these solutions, but feel locked out of actually getting in past the people who are obviously not up for the challenge.
I use Claude and ChatGPT daily and I have my own list in a spreadsheet that I keep an eye on because things are changing rapidly. I am going to try a few more now based on your reviews - in particular, IO 💛🙌
Finally, I am part of a pretty large engineering and product group, and we are discussing these solutions constantly and reviewing. I am working on my own product based on AI on the side as well as writing books and music videos and content trying these solutions daily. It’s a lot of fun!
I appreciate the reverse-ordering for the article. There was no way I would make it to D level if it was at the end of the page lol.