I Tried Every AI Productivity and Coding Tool, These 7 Will Save You the Most Time (Nov 2025)

Hey everyone. All right, so today I want to share my ranking of 22 popular AI tools that help you save time and get worked up. Out of the 22 tools I tested here, there are only seven tools that I put in the S tier that I use every day to work and code. And a few of these tools I use even more than Chad GPT. Okay, so let's get right into it by ranking the AI tools for everyday work. We'll start with the worst tool

and that is Gemini for Google Workspace. It can't really do much of anything. You know, it can't summarize comment threads or edit documents at this point. I think Chat GPT and Claude have better Google Workspace integrations. And I think this is the worst AI product that I've tried because it just has so much wasted potential. Let's just take a quick look. So here I have a Google doc for the AI tools landscape and let's open Gemini and let's see what can Gemini do and you know it gives you a huge list of stuff here that you can do. So let's say why don't you edit the title to AI coding tools instead and lo and behold it cannot directly edit the title. Why don't you summarize the comments and it can't do that either. It can summarize a document, but you know, that's pretty standard. Anyone can just do that. So, yeah. So, I think this is probably the worst AI tool I've tried, unfortunately. And maybe Google should put the Gemini team in charge of workspace since they seem to be shipping just fine. All right. Next, another tool in Dtier is Zoom AI. You know, Zoom AI takes meeting notes fine, but the notes are incredibly long and just a chore to read. I have to use chat GBT just to summarize Zoom AI summary to get the takeaways. So, Zoom does let you customize the template, but of course, they neglected to let people include their own prompts and templates in their product. So, overall, this product is a far cry from an AI meeting tool that I'll talk more about in the S tier. All right, moving on. In the C tier, Grammarly just keeps changing what I actually mean to say. I was a loyal Grammarly subscriber for years, but I cancelled because I kept on suggesting changes that completely change the meaning of my sentences like this example. Now I just ask Chad GPT or Claude to fix spelling and grammar in my draft and bold any edits that you make to update my grammar instead. Now Shashir is a leader who I respect a lot and he recently took over Grammarly and he renamed the company to Superhum which also includes Superhum's email service. So I'm optimistic that this tool will improve fast but for now I can't really find a use out of this tool. Okay. So next let's move on to the beat here. The first tool here is glean which is very useful for internal knowledge search. You know it's very good for queries like summarize all documents related to topic and look up recent information about this product or tell me this team's goals. Right? But on the flip side, I don't find Glean's AI agents as flexible for workflow automation as a Zapier or NAN. And I'll talk about workflow automation in more detail in a future video. So overall, it's really solid workhorse if you're trying to look up and summarize information at your company. All right, so the next tool in the B tier is Gamma, which generates

slide decks in under 60 seconds. All you have to do is paste your notes and then customize the amount of text, theme, and image style that you want to start. So, let me show you a real example. So, let's open up grammar. And here I give you a few options. So, let's say we want to paste in text. And I have my written blog post for this video right here. So, I'm just going to copy and paste this whole blog post into Gamma. All right. So, now let's click next. And uh here you see Gamma has broken out my blog post card by card. And for slide decks, I always like to have minimal text. So let's do that. And let's pick this visual style. And for image art, why don't we pick, let's see, why don't we pick line art right here. Okay. And let's just go ahead and hit generate and see what it comes up with. All right. So you can see here that gamma is generating my slide deck. And this is the title slide. It's created a line art image and it's going to go ahead and keep going for Dtier tools, Ctier tools, and it's going to get to the AI coding one next. So, you know, this is a pretty impressive AI product, right? I mean, the fact that it can just generate these slides from my outline that I didn't edit at all is pretty impressive. And of course, I can go in here and change the text and, you know, even update the images a little bit. But you can see here that the AI generation, you know, it just kind of feels a little bit too AI generated. I think it still has a little bit too much text. The images that it generates aren't really relevant to the contents of the post. So that's why still put in the B tier, but I'm really excited to see this product evolve because as a product manager, I really hate making slides. And if Gamma can support better images, if they can support importing templates from Google Slides or PowerPoint, I think this can easily be an A tier tool. All right, so now we enter the A tier, which is tools that are actually very good. And the first one I want to talk about is linear. So linear makes it easy for teams to assign work to AI agents. Linear started as a issue tracking tool but has quickly evolved into the best platform for teams to coordinate with AI agents like cursor and GitHub copilot and it's actually pretty amazing like so for example here's a tweet from Linear what you can do is even in Slack you can just tag linear agent on a ticket and then it will go ahead and start trying to fix the bug for you and then at some point it's going to submit a pull request back for you to review. So this feels like the future, right? Being able to assign bugs and other work that you don't want to do to AI agents and then saving the more interesting work for humans. So yeah, linear is a very high craft product. I'm excited to see them expand in this direction. Okay, next we have notebook LM. And notebook LM is really good at making your boring research notes extremely engaging. So, if you have a lot of documents and files that you just can't be bothered to read, just upload them to Notebook LM to get an engaging podcast conversation or a short video overview summary. So, for example, I copied the whole transcript of this video into Notebook LM and I asked it to make a video overview, right? So, let's take a look at what I came up with. All right, so let's go ahead and play this. just take a scrub through what they have here. This stuff is pretty impressive, right? Like these are like custom illustrated graphics. Personally, I think the slides he notebook makes are much more interesting than slides like Gamma makes. So if you want to get a quick video overview, mind map or audio overview, Notebook LM is really a game changer. So I definitely recommend using this to learn material to ingest a lot of documents, a lot of sources that you want to read and get an audio overview or video overview instead. Okay, the next tool in the A tier is Super Whisper. And Super Whisper, probably haven't heard of it, but it's perfect for local voice dictation while going on walks. Now, Super Whisper is a voice dictation app, and it supports local models for privacy, and it is the only voice tool that lets you record for over 10 minutes at a time. So, I have to go on a walk in my neighborhood park picture here. And I usually pull up Super Whisper and I dictate a first draft of a document to the app. I then use Claude to clean up the draft afterwards using my prompt. So, Super Whisper is pretty amazing if you want to just like go on a walk, clear your head, and dictate to the app. Here is an example of a purely voice dictated notes that I sent to myself. It's a first draft of a blog post that I'll write. and then again I'll use claw to clean it up afterwards. Now the only drawback of super whisper is that in terms of transcription accuracy it's not as good as the s tier tool that I'll cover next

and that tool is whisperflow. Now if you try only one tool from this video I recommend that being whisperflow. It is a voice tool that I use even more than chat GPT and voice dictating accurate notes with a single button press. It's just a gamecher for my productivity. I also find Whisper Flow more accurate than other voice tools as it auto corrects mistakes and can even format your notes into bullets on the phone. It also supports mobile keyboard integration which means that I can dictate accurate notes and feedback from any app that I use. All right. So, let me kind of show you a demo of Whisper Flow. Okay. So, this is Whisper Flow and let me show you a quick demo. So, all I have to do is press down the key and let's go ahead. So, I am going shopping and I want to add three items to my shopping list. First, I want to add tomatoes. Second, I want to add milk. And third, I want to add bread. Uh, and you can see here other than this part, let's cut this out. um is actually formatted my voice dictation into bullets, which you know, as far as I know, no other voice tool can do. And another really fun feature of Whisper Flow is that you can actually include snippets. So, let's say I'm giving a talk and I want to insert my own bio. So, all I have to do is hold down the key again and say Peter's bio. And you can see instead of saying Peter's bio, it pasted my actual bio. Right? This is a very humorous bio. It's not a real bio, but you know, this is a bio. And that is because I put in a snippet right here that whenever I say Peter's bio, it past my real bio. So yeah, so I think Whisper Flow is really a game changer for me. And you don't have to use Whisper Flow if you really want to, but you should definitely be using some sort of voice dictation. Okay, so let's go on. And the next tool in the S tier is Granola. Now, Granola is an AI meeting nodes tool that is just like much better than Zoom AI or other AI meeting companions out there. And unlike Zoom's verbals summaries, I can edit Granola's prompt to get nearperfect notes each time. So, here's my prompt that I use. Generally, after a meeting, I like to just have a few takeaways with both stems and then also have a few next steps, right? So this way I can basically get the meeting notes from Granola and then just copy and paste that directly into an email or a Slack update without having to process it through chat GPT or some other AI. So that is pretty amazing by itself. But Granola is also building useful features like chat so that I can actually ask Granola, hey, you know, summarize my meetings this week or give me some post backtoback meeting therapy. If you think about it, the most useful information in a company is the recent meetings that you've had, right? And Granola has this information for better, for worse. So yeah, after you have a couple of meetings with Granola, you can definitely ask it to share some insights or share some takeaways across your meetings. It is extremely useful. I think this tool alone probably saves me at least four or five hours a week because, you know, I don't have to take meeting notes anymore. Okay, so that's Granola. Now, the next tool in the S

tier is Notion. And Notion, I think, has really built the best AI agents for everyday work. It is the exact opposite of Google Workspace AI in that it actually does more than you think it can do. And the killer feature is being able to make bulk updates to Notion database entries in one shot. And it can also search Slack better than Slack's native AI features and do a lot more. All right, so let's take a quick tour of Ocean AI. So here I have my prompt library, right, that I use. It has a lot of great prompts. And if you want to get this prompt library, you can check out the link in my newsletter and subscribe. But let's actually pull up Notion AI here. It's a little person with a cat here and you can personalize it, change the hat that it has. But let's actually do something here. So let's actually ask it to can you remove the date field from each entry in the database. Okay. So if I had to do this manually, I have to click into each entry and try to remove the date field, right? But because I have notion AI, hopefully it can just remove all the date fields for me. Okay, so it's asking for clarification. Okay, so I've clarified my request and let's see if it actually does a job or not. All right, there you go. So, looks like it's actually removed the date field from all the entries in one shot. And just for fun, let's actually make a new notion page. And now, let's say I have a new notion page. Can you add a bunch of sample blocks to demonstrate with fun what you can do? And you know, unlike Google Workspace AI, Notion AI can actually edit Notion documents. Let's take a look at what it comes up with. So, it came up with block sampler. It's created headings, lists, to-do lists. It's created expandable blocks, tables, images, even videos. I mean, it's done a lot. So, yeah, notion AI is actually quite incredible. I think it actually addresses the biggest complaint I have of Notion, which is it just like a lot of work to manually create these blocks. But with Notion AI, it makes it incredibly seamless. All right. So, if you haven't tried Notion AI, I definitely recommend giving it a try. Pretty impressed by it. And I also have an interview with Notion AI leads Ashe and Brian who gave me a live demo. So, you check out the link in the description for that. Okay. So, the last AI tool I want to cover in the work category is Perplexity. And we're all familiar with Perplexity. I use it more than Google search to be honest with you because it is lightning fast and I use it for things like if I want to go somewhere for example this place in Mexico in December, I use it to look up any kind of familyfriendly activities or things going on or if I want to look up just quick facts. For example, I look up what are some stats about Perplexity new comet browser. Perplexi is the fastest way to do this lookups because it can find recent information incredibly fast. It generates much faster than chat GBT, right? And you can keep going back and forth with it. So to be honest with you, I use Perplexity probably multiple times a day, mostly to look up toate recent information about different things I'm working on or vacations I'm planning or even local familyfriendly events. So yeah, so I am a loyal perplexity daily active user. The last thing is they also have a pretty generous pro plan. With Perplexity, you can get five free pro queries per day, which is more than enough for the average user. So, you probably don't even have to pay for Perplexity to get value out of it. All

right, so now let's cover my favorite AI tools for coding, which is arguably the most competitive AI market out there. Now, we all know about cursor, but some of my picks here in the higher tiers might surprise you. So before we get started, I want to talk about my highlevel perspective of the AI coding space. You know, I think companies are really only willing to pay for two AI coding use cases. One is to help engineers ship production code and this means using AI to refactor code, handle migrations, add features, and fix bugs. And two is to help non-engineers prototype zero to one. So this means helping designers, PMs, marketers, and founders build apps, websites, and internal tools from scratch. Now, my hot take is that the ship production code use case has much clearer ROI than prototyping 0ero to one. With AI coding, the most value comes from solving the boring, high friction workflows that engineers don't want to deal with manually. I'm talking about stuff like migrations, refactors, code reviews, and documentation. And in contrast to prototyping, the ROI is just much harder to measure, right? Like how much faster can you validate an idea with AI instead of using Figma. And when every prototyping tool produces the same output, you know, what's the real differentiation? So prototyping is still valuable, but production code I think is the bigger market right now. So with this in mind, let's start with my beer tier for AI coding tools. Okay, let's kick off the B tier. We have Windserve and Devon which are still very much in the race but just don't seem very differentiated against the competition. You know most engineers I talked to now default to cursor for their IDE and either codeex cloud or cursor for their agent. Now Windsorf did just launch their new sui 1. 5 fast coding model which seems to be pretty comparable to cursor's new model. And it's actually kind of funny because like both of these fast coding models are actually trained off Chinese foundation models. But I think that's like another topic. So Windserve has this comparison list between Winsurf against cursor. And overall I just don't think Windsorf is that differentiated except that it's $5 cheaper a month. Like for example best for enterprises and large code bases and best for small teams and solo developers. I just don't think this is true. I I think cursor has incredible adoption in the enterprise space and Devon which is the AI coding agent. It also seems a little bit undifferiated when openi codeex and some of these other alternatives can also code async for hours at a time as well. So overall I'm really glad that windsurf and devon are still in a race but I put them in the B tier. All right this next pick

might be a little bit controversial. Lovable, I think, is the most beginner friendly AI coding tool out there, but it struggles with complex apps. You know, Lovable is very good for building personal websites and simple apps as seen on their community showcase. Let's pull this up. So, if you go down here to their community and let's sort by popular, the most popular tools are I mean there's much of a crypto dashboards, but other than that, the most popular tools are kind of websites for hotels. It's beautiful, right? And there's other tools around here, you know, internal dashboards, things to prototype games. So, I think Lovable is very good for websites, simple apps, and maybe some internal dashboards. But if you want to build full stack apps with backends and databases, with authentication, with other kinds of features, you're probably better off using replicate, cursor, cloud, or codeex. Now, similarly, both I will also put in the B tier. I actually really enjoy using Bolt and it seems like they're kind of pivoting. They're pivoting to be a web platform for the top-of-the-line AI coding agents. So, Bolt V2 now supports clock code in a web UX that's arguably more approachable than an IDE or a command line tool. And their web containers tech also makes hosting and deployment really easy. Okay, so let's say let's build Pitus. And if you look at this, Bolt does thinking first. It very much kind of feels like you're using claw code in this kind of interface. Okay. Now, the drawback of Bolt is that codeex isn't quite supported yet from OpenAI and the clock code team also shipped their own web interface which kind of competes directly with this. So, I'm pretty optimizer for both, but I feel like being a web platform for top-of-theline agents, you're competing directly against the agents, too, right? So, it's unclear how this will all play out. Okay, so let's keep going. And the next one I put in the B tier is Figma make. And Figma make the thing I want to highlight is that recently they launched design system support, but there are still some quality gaps that remain. So Figma's new make kits now let you import your design systems React components and CSS files into Figma make which is arguably one of the biggest unlocks for AI prototyping because all the other AI prototyping tools there's a very high chance that this produce like this purple AI slot the designs when what you really want is designs that actually match your company's design system right so this is probably the biggest unlock however the feature is still early and as these users report some buggy output export issues. If Figma actually makes as design system feature very good, I think AI prototyping space is Figma's market to lose because everyone already uses Figma for design. So having make take off just seems natural. All right, so let's see how the design system feature plays out. Okay, so now we move

on to the A tier of AI coding tools. And the first one I want to talk about is replet agent which is powerful for both beginners and experts. Replet agent is generally pretty great. It can run code, it can write tests and it can even screenshot its builds to iterate by itself. So combined with Replet's full stack platform which offers backend and database inhouse, I'm really not surprised that Replet is still gaining traction inside companies. So here's an example of me using replet agent to build a trivia game. So this is a trivia challenge game. I can pick single player or I can pick multiplayer. I can take trivia challenges and I can, you know, answer questions. And if you look at the chat history here, it's doing a lot. It's like testing its own product. It is looking at its own web browser and console log to see if there's any bugs. Like these are features that not a lot of other AI coding tools have. So basically replet agent can often oneshot simple apps. I recommend actually beginners if you want to build anything slightly complex to go to replet first. All right. So next another tool in the A tier is factory droid which has really impressive benchmark performance. You probably have never even heard of factory joys right but they are second only to claude in terminal bench performance. You can see here that for terminal bench cla is first and then factory is right afterwards. So many companies are already using joys for refactors, migrations and other engineering tasks. So factory joy is definitely very impressive. All right

cool. So now we enter the S tier of AI coding tools. And the first one I want to talk about is claw code. Claw code pioneered features that competitors are now copying. I'm talking about features like plan mode, slash commands, and many other tools. They were all pioneered by cat Boris and the clock code team before competitors started copying them. And as a PM, what I really love about clock code is how it's useful for things beyond coding. I use it for tasks like research, news curation, and metric tracking. I had to put clock in S tier because it is a pioneer in the space and it continues to innovate. Now, this next tool is a direct competitor to Clockode, which is OpenAI Codeex. And OpenAI codees excels at planning complex changes with really surgical precision. So many of my AI native PM and developer friends now default to codeex to plan complex features and to fix gnarly bugs. So check out this quote from Colin, right? He had a small bug that he was working with cloud on and he gave it to Codeex and it thought for a while but it managed to fix the bug in only six lines. For developers, being able to fix stuff in a few lines so that you can actually review it and confirm that it actually works is kind of a gamecher. So, codeex I will say is probably top of the line. It's currently probably the best AI coding tool for planning and fixing gnarly issues. And last but not least, we have cursor. I recently introduced a lightning fast composer model that complements their best-in-class IDE. You know, cursor already offers the best coding ID, but many people, including me, were using cloud code and codeex instead of cursor's native agent in cursor, which is probably bad for the cursor team. And this change with cursor 2. 0's new super fast composer model. You know, it may not plan as deeply as codeex or claude, but it's just much faster for quick iterations. All right, so actually, let me open cursor and show it to you. So, cursor actually has a new agent mode that looks pretty suspiciously like lovable and built and some of these other web-based tools. It just has the agent and the output. But let's go ahead and ask it to build. So, composer one alpha. That's the model that we picked. Let's go ahead and do it. Okay. So, see how fast this is. It's like literally like, you know, generating all this code at lightning speed at a speed that I can't even read. Okay. So, that probably took like honestly like maybe two minutes to generate. And let's actually open our browser and see if the game actually works. And there you go. It looks like they built it's working Tetris game in literally just two minutes. This is pretty amazing. It's lightning fast. All right. So Rio is the designer for cursor. So it's pretty biased, but he is saying that the best way to code right now is to use cursor, use composer one for live coding changes and use GPD5 high or codeex for planning. And I tend to agree with him. I think this is the best combination right now. That's why cursor, Codeex, and Clock Hold are all in the S tier. So, there you have it. Out of 22 AI apps that I just covered, here are the seven that are S tier for me for work and coding, right? So, starting with Whisper Flow, which I use for AI voice dictation, it is by far the most accurate. Granola for AI meeting notes is amazing if you give it your personalized template. Notion AI is just an amazing agent and I had to put in the S tier because it's just so much better than Google Workspace. Um, and Perplexity that I use every day to do up-to-date information searches. And for coding, I think the S tier is really three tools. So, Clockode, which is the pioneer, the innovator in the space in my opinion, Codeex, which is right now the best coding model for planning and fixing gnarly bugs and issues. and cursor, which is a really great IDE. And they also have a lightning fast coding model that's brand new. If you enjoyed this video, please like and subscribe. And in my next video, I'll rank my favorite AI tools for workflows, agents, browsers, creativity, and even just for having fun. So, be sure to like and subscribe for more practical, nononsense AI tutorials like this. Thanks so much. See you around.