Why Are These 19 AI Apps Still Free?

Matt Wolfe walks through 19 free AI apps: coding tools, image and video generators, productivity agents, and gaming simulators. He shows what each one does and warns that free access won't last forever, so you should try them now while you can.
We've got Minesweeper. Does it actually work? Oh my gosh, yeah. Minesweeper actually works inside of this Windows XP simulator. As somebody who looks at and tries out AI tools every single day, I am constantly on the hunt for the absolute best free tools that people can use right now. One, because well, they're awesome and two, because who knows how long some of these are going to be free for. Some of these companies have a tendency to let you try them for free and then once you're hooked on using them, well, then they start to charge you. That's pretty, pretty clever, dad. So I want to show them off now so that you have the opportunity to try them out while they're still free. And let's start with the free AI tool that I probably use more than any of the rest and that's Google's AI studio. You can find it over at aistudio.google.com and well, I'm not going to go too deep into this one because I literally made an entire breakdown video of all of the cool things that you could do for free with Google AI Studio. So definitely check that video out if you haven't already in that video. I use their free AI code generation tool to create this emoji fusion mastered game where you can combine elements and it turns them into new elements like Celestial Fluff. I created this Alien Dodger game which works a lot like Frogger, but you're an alien. And I even completely cloned one of my favorite apps, the Feedly app, all using Google's AI Studio coding tool. Inside of the free AI Studio, you can also do text to speech, you can generate images, you can create music in real time, and you can play around with pretty much any one of Google's AI models that are available out there. Another really cool feature that you can do again totally free inside of AI Studio is that you can share your screen or share your webcam while having a conversation with the AI and it will actually walk you through how to do things. So for example, I could look into my YouTube analytics here and get advice from the AI based on my YouTube analytics. What types of videos should I focus on if I really want to grow the amount of subscribers my channel has?
Based on your analytics, the video that brought the most subscribers is the one about 50 plus insane new ways to use the new ChatGPT. Also, the AI news updates seem to attract quite a few subscribers as well. Focusing on similar topics like AI news and practical uses of specific AI tools could help you grow your subscriber base.
It's super cool, and I could come back and see exactly what the AI saw. I could listen back to what I asked it, and I can listen back to the response that it gave me. And since we're talking about Google, I need to show off another Google tool that's totally free to use that I absolutely love, and that is NotebookLM. This is a tool where you can upload a bunch of documents or website URLs, or even plug in videos and then ask it questions about those videos. It'll make mind maps about those videos. But the feature that I like the most is the ability to actually create a podcast around the information that you feed it. So recently I got a new camera called the Fujifilm X106, and I wanted to learn how to use it. So I went and found a handful of resources all about it. Plug those resources into NotebookLM and we can see that it created a mind map for us around it right here. And I can dive deeper and deeper into this camera's features and, you know, technical specs and things like that by digging deeper and deeper into this mind map that it generated for me. But it also generated this podcast that essentially explains more about the camera. So I can listen to a 20 minute podcast and get a deeper understanding of this new camera that I just purchased. Have you ever looked at a photograph.
Maybe one taken with, you know, an incredible camera like the Fujifilm X100VI?
And the truth is, it's probably not the camera.
It's more about really knowing how to unlock all that potential packed inside it.
And I use this Notebook LM tool all of the time. It's so amazing when I'm trying to dive deeper on a specific topic to just throw a ton of resources in here, you simply click this add button. You can throw in PDFs, text files, markdown files, MP3s, PNGs, JPEGs, YouTube links, Google Drive files, or just copy and paste text. And once it's in there, you get the mind map, you get the audio podcast, and you can have chat conversations with all that information and ask things like what essential settings and techniques optimize the X106 for diverse shooting scenarios. And we get a response with a lot of details and the exact source. If I click into one of these links, it'll actually pull the exact document it found it from and highlight the section that it actually got that information in. This is super powerful and still blows my mind that they just let anyone use this for free and we can't Talk about Google without talking about Google's VO3. VO3 is their new video generation model, which is really great. It's really realistic. It can generate audio. All of the memes, all of the crazy viral AI videos you're seeing lately were probably generated with VO3. Now, VO3, not free to use Google, actually charges quite a bit to use it. However, there is a trick that's working right now. Can't guarantee this will work forever, but right now you can actually generate videos for free with VO3 using Perplexity's Xbot. So if you go over to X and you submit a prompt on X where you tag Ask Perplexity, you can actually give it a prompt and it will make that video using VO3. Now, it's using VO3's fast model, so it's not their highest quality model, but it's still pretty dang good. So you can see here I went, ask Perplexity, make me a video of a talking cat that's sitting next to a taco and saying, subscribe to Matt Wolf. And here was the result. Subscribe to Matt Wolfie. Now, it was supposed to be the cat saying that, but it's AI. Maybe a couple more attempts and it would have gotten it right. But that is a way that you can go play with VO3 right now for free. I mean, depending on when you're seeing this video. And since we're talking about AI video, let's talk about cling. Cling is probably one of the better AI video generators available on the market right now. And they do have a free tier where you can generate videos for free. Now, they take a little bit. They do prioritize paying users, but you can generate videos for free with this tool. If I go to clingai.com and then come up and click on Create, it will jump me into the Cling platform where I can generate videos. Now, if we look at their pricing here, you can see that for $0, you get credits by logging in. So the more often you log in, the more credits you get. And those credits can be used to generate videos. I logged into a free account. You can see this is not a paid plan. It's telling me to upgrade. And it gave me 166 cred to play with. I could give it an image like this one with a woman and a cat and give it a prompt. Like a woman gets closer to a cat and the cat gently sniffs her. A gentle breeze brushes her hair, the cat's fur and the flowers in the background as clouds Drift away and the camera zooms in. And you do have options for other camera effects as well. It will also add sound effects to the videos, very similar to V3. And we can see it's going to use 20 credits. Again, I've got 166 to play with. Let's go ahead and generate and we get a video that looks like this. It can also do text to video. So if I want to use my a monkey on roller skates prompt, I get something like this. So again, that's cling AI. You can find it@clingai.com the video generations aren't the fastest if you're on the free plan, but I mean, it's pretty good for free. Now let's talk about image generation. My favorite image generation platform by far is Leonardo AI. Now, I'm a little biased. I do have some equity in the company, but I have equity because it was a company I loved. In my opinion, they have the best platform and user interface for generating AI images. And their new Lucid Realism model is probably my current favorite AI image generation model. Now, I do have an upgraded account on Leonardo, but I'm going to log out and log in with a free account so I can show you what you actually can do for free in Leonardo. So this is a brand new Leonardo account here. I'm not a paid user. This is completely free. You can see that it's telling me I'm on the free plan and I need to upgrade if I want priority generation. But I have access to these models, including the GPT1 image generation model here, the Flux Context model, and my current favorite Lucid Realism, which is an exclusive model just inside of Leonardo. I'm going to have it generate a random prompt for me, see what it comes up with. And it's going to cost me 66 credits to generate this. You can see I've got 150 for free right now and that is refreshed every day. Also, keep in mind I can use less credits by making the images smaller or by switching to different models. Different models have different amounts of credits they use. Like the Flux Snell model will only cost me 8 credits to generate images with it. The Leonardo Phoenix model, only 20 credits to generate four images. But again, I really like this Lucid Realism model. So let's go ahead and generate the suggested prompt here. And you can see the quality of images that it generated here. They look pretty good. I feel like it's hard to get a real feel for the style when it's more of this like anime style. So let's try to get something a little more realistic. I'll just do a man stands on the sidewalk in front of a busy city skyline. We'll go ahead and generate. It's actually going to improve that prompt for me. You can see how it just changed the prompt to give it a little bit more detail and we get some solid, pretty realistic looking images. You can even do things like edit the images similar to what you can do in OpenAI. So if I click into one of these images and say put a baseball cap on the man and click Generate, we can see that it generated an image. It's the same image, but now it put a baseball cap on the man. Now I'm pretty much out of credits for the day and can come back tomorrow and use more credits. Or next time I can try different models that also use less credits. Here's another fun one that I recently came across called Things. It's actually spelled T H I I N G S and it's a whole bunch of AI generated icons for pretty much anything you can imagine. And if it doesn't have one that you want, you can actually request that they generate it. Now you can generate your own things, but it costs credits and they don't seem to have any sort of free allotment of credits. But you can suggest a thing and if it's not in their database already, they'll create that thing for you. And chances are whatever you're looking for an image of, it's probably gonna have it. Like if I search a wolf, there's definitely a handful of options available to me. If I search a monkey, we even got monkeys smoking cigarettes. We got tacos, various animals, various vehicles, like drones that look like the DJI drones. I mean, there's probably a good chance you're going to be able to find the icon for whatever it is you're trying to do in here. And it's probably already been created. Pretty cool resource that I thought would be fun to share. Next up, let's talk about Whisk. This is another one out of Google that's free to use and it allows you to sort of combine concepts into images. And I found it to be really fun. Let's go ahead and click Enter Tool. This one you can find over at Labs, Google FX Tools, Whisk. Let's go ahead and enter the tool here. And if we open up the left sidebar, we have various elements that we can tweak, like the subject, the scene, and the style. So for the subject, I Like to upload an image, I'll upload an image of myself here. And then under style you can either enter a text, upload another image, or we can do something random like here's an anime looking image and then for scene, let's just put in space. So we've got me as the subject, sort of space is the scene, and then in an anime style and I can go ahead and click Generate and it mixes all those three together to make an image of me in space in an anime style. They've also got some fun presets. So if I click on the little hamburger menu here, there's some presets like a sticker, enamel pen, plushie card, chocolate box, capsule toy, bento box. I've always thought the plushie was kind of fun. Give it an image of myself. Let's get rid of the background scene and then let's use style and just use those two and click Generate. And now we have a plushie me. Now there are buttons here to animate these if you want, but I believe this animate button's only gonna pop up if you're on one of the paid plans. But the ability to sort of generate images by combining images that you can do for free. One of these days I really wanna get one of these plushies made. I think it'd be really fun. Next up, I need to tell you about Glyph. Glyph is this like AI workflow builder where you could sort of tie together different AI tools to get a different outcome and then share those glyphs so other people can use them. And you know what, it's probably best that I just show you. So if I go to Glyph app, that's G Lif app, I can click on Explore and we can see some of the other glyphs that people have made. So for example, someone made a YouTube thumbnail maker I could click into. Here it asks for a prompt. The example, an Android robot smiling and showing a phone screen which says Glyph app. Sure, let's run this glyph. It's going to use 4.12 credits. You actually get an allotment of credits every single day. I actually currently have a thousand credits and I've never paid for them. You just get 10 a day and they seem to stack up. And we can see here, this is the workflow for this glyph. We gave it a prompt, it's actually enhancing the prompt. It's setting the width, the height, the steps, and then it's doing a bunch of comfy UI magic stuff in the background. Using AI to give us our final YouTube thumbnail. Now, not my favorite YouTube thumbnail I ever came across, but not bad for just giving it a single prompt. Let's check out some of the other glyphs that people have made. Like here's a glyph called Spider Verse. It hallucinating the style of the Spider verse movies. So let's put, I don't know, let's just put Batman, run this glyph. And I guess this is what Batman would look like in the spider verse. Probably pretty accurate. Looks good to me. But where glyph is really powerful is building your own workflow. So let's go ahead and click on build here and you can see you get like these glyph blocks. So let's go ahead and start building with glyph blocks. We want to start with a text input. This is where somebody could enter a prompt. Let's reword this to enter your prompt. And then we want it to take that prompt and maybe use a text generator to improve the prompt. We could select any of these models here, which is pretty much every available model. Let's use GPT4O and we can give it a prompt like improve the following image prompt so that it has brilliant colors and beautiful lighting. Also add a wolf somewhere in the image prompt. There should be a wolf in the background of the image. And then we'll go ahead and give it the input. Then I'll go prompt. And it knows that this is the prompt to improve. So now you enter an image prompt. It'll run through GPT4O, improve the prompt, add a wolf to the prompt. And now we want it to actually generate an image. So we'll use an image generator. And then for our image, we're going to pull in our text here. So we'll use text 1 and then we can select what type of image generator we want to use. I'm going to use flux Pro version 1.1 Ultra and let's make it 16 by 9. And now we can test this glyph. Let's just do Batman again, right? So my prompt is just Batman, but this is going to make it colorful and add a wolf to the image with every single image. Let's run this glyph and just by giving it the prompt, Batman, we got a beautiful colorful scene with Batman and of course a wolf somewhere in the image. And I can give it any prompt I want and it will give it those additional details to add the brilliant color and to add the wolf. So I can call this glyph Give me a wolf. Then I could publish this glyph and anybody can go use it. Now. Now, if I go back to my Explore page and look for new, you can see my Give me a wolf glyph is available for people to use. But this isn't just for image generation. You can have it generate videos if you know what you're doing. With Comfy ui, you can build out entire Comfy UI workflows. You can have it generate aud, you can have it take an image and take it down to text. You can make audio to text, you could extract JSON, you can make HTML, you can give it a video input, you can have it generate videos, all sorts of workflows, and you can make these workflows as long and as complicated as you want. And again, you get a certain amount of free credits every day and you can just play around with this stuff. Next up, let's talk about 11 labs. This is probably the best audio generator available, especially for text to audio, and you can do a lot with it for free. Now, I do have a paid plan, but I'm going to log into a free plan where I'm not paid on this account and show you what we can do for free. Once I'm logged in, I could use their text to speech generator where they have tons of voices that I can use, like Alice. Just trust yourself, Aria. There is no greater harm, Bill, as we are liberated. Brian, the thing always happens that you really believe in. Let's go ahead and use Aria and generate some speech. Subscribe to Matt Wolf on YouTube and check out Future Tools IO. Not bad. Again, this is on a free plan. You can also use the Voice Changer where you upload an audio file and have it change your voice to a selected voice and you can generate sound effects. So if I need a sound effect of a car crashing through a glass window, I can generate that and it gives me four versions to choose from. And there's a ton of other features in 11 labs that you can use, but some of them, once you start to use too much credits, you will have to pay, but you can still do quite a bit for free. Since we're talking about audio, let's talk about Suno. SUNO is a tool that creates AI generated music, and in my opinion, it creates some of the best AI generated music. Now, Suno does have both free and paid plans. If we scroll down here, we can see that on the free plan you get access to version 3.5 so you don't get the absolute latest state of the art music generation model, but 3.5 still pretty dang good. So when I'm logged into SUNO here, I can literally just describe the type of song I want. For example, a pop punk anthem about subscribing to Matt Wolf on YouTube. I'm going to go ahead and set this to version 3.5 just so you can get an example of the type of music it will generate on the free plan. And let's go ahead and create. And after just a few seconds, we've got two options to choose from. Subscribe to my full teeth they see you'll succeed at lightning speed. Smash that bell. Hear the one indeed. And here's the other song. All right, so far we've talked about various creative pursuits that you can use AI for. Well, now let's get into productivity because there are a ton of free AI tools that will make your life so much easier. One of them for me is Perplexity. I love Perplexity. So much, in fact, that I did an entire video about all of the cool ways that you can use Perplexity. So if you haven't checked out that video, make sure you check that one out. Like most of the tools on this list, this does have a free plan and a paid plan. I personally do subscribe to the paid plan and love it it, but I'm logged in on a free account here. This is not a paid plan to show you that you can still get great results on the free plan. For me, Perplexity has pretty much entirely replaced Google Search. Anything that I would have searched on Google in the past, I just search Perplexity now instead. And I literally just find myself asking it random questions all day long. For example, I was flipping through the channels one day and there was a Texas A and M football game on. And one of my kids said, what does Texas A and M mean? And I went, I don't know, let me ask Perplexity. And with a quick little prompt, well, it stands for agriculture and mechanical. I really like Perplexity because it essentially goes and searches the web for you. And then once it searched the web, it goes and summarizes the results and then even cites its sources. So if you're kind of uncertain with the information that it gave back, you could always go and check the source of where it got that information. This was obviously not the most complex question to ask Perplexity, but it works well for demo purposes. Next up, I want to show you Mind Studio, another tool that I've used so much that I've literally made multiple other videos about how you can use Mine Studio. But if you're not familiar, there's really two main ways to use Mine Studio. You can build these visual workflows in here, kind of similar to what we did with Glyph, but I feel like it's way more powerful and has way more various integrations that you can tie into it. I could do things like create a user input here. This would be where you would give it like a prompt. And then once you give it a prompt, there are so many different things you can do from here. You can have it generate an image, generate a video, a chart, create a podcast, scrape a URL. Now, some of these features, I believe, are only available on the paid plan, but a lot of this you can still build out for free. It can Fetch info from YouTube videos and YouTube comments and scrape Instagram mentions and search Blue sky posts and update Google sheets. It's similar to what you might get out of a make.com or an N8N and honestly, really, really powerful when you build some workflows. But instead of showing you building a workflow, let me just quickly show you inside of an existing workflow. You can see that the ones that say Pro Agent, those will require a paid plan to actually use. So let's pick one that's not. So we've got find product alternatives side by side product and service comparisons. Let's click into this one here, make a copy, and we can see what this workflow looks like. This is one of the workflows that's available for free. So you have a user input. It's going to ask you for an intro, a name, and a URL. It's then going to scrape that URL. It's going to analyze the product using llama 3.170 B. Then it's going to find and analyze competitors using Perplexity. Then it's going to generate a side panel report. Then it's going to display the report. And here's what that looks like in practice. If I click Publish and then open the AI agent, it's going to ask for a product name and the product homepage URL. So let's go ahead and put Future Tools and the homepage URL futuretools IO, we'll click next and then it goes and runs those processes in the background. And after roughly a minute or so, it came back with this report that it generated for us. Future Tools, a platform that collects and organizes AI tools to Help users become superhuman. It lists off the key features of the website and then it shows some competitor websites like Toptools, AI, similar web favored, and there's an AI for that, also sharing strengths and weaknesses of each of these competitors. But mindstudio also has a Chrome extension that's pretty cool. You can see it adds this little icon on the right sidebar, and it allows me to access various agents directly inside of my sidebar. So to show what this extension can do, I opened up a TechCrunch article here all about Sam Altman and the New York Times. I can use one of these quick bots here to either give a TLDR or an executive summary. Let's go ahead and run the TLDR bot and we can see that it gave me a quick summary of what this article is all about in my little TLDR summary here. Again, this is just scratching the surface, but there's all sorts of little bots that you can run on various web pages to get extra information or to automate processes based on the content of the page and things like that. It's actually really, really cool. Here's another one super, super simple concept. It's called URL to any you give it a URL, it'll give you back either a markdown file or an HTML file, or a PDF or an image or a JSON or a QR code or whatever you want from that initial URL. So let's go ahead and go back to this TechCrunch article we were just looking at. I will copy this article URL and let's say I want a PDF version of this URL. I'll do URL to PDF, plug in my URL and click convert within about 2 seconds. I now have a PDF version of this article right here that I can download and use. However, I want just something to bookmark that you'll probably find handy from time to time. All right, now let's get into even more of the like, AI agent kind of stuff. This is a brand new one that just recently rolled out called 11ai. You can find it over at 11ai. It's actually created by 11 Labs, but it's designed to create a personal voice assistant similar to like Siri or Alexa that ties into a lot of the products and online services you use. And if we click this little learn More Button Here, is 11AI free to use? Yes, all calls through 11AI are free during the alpha period. So as of right now, this is free to use. But again, this may not be free. Forever. So we can go ahead and name our assistant. We'll just leave 11 as the name. We can pick a voice that we want to use. Let's do Burt Reynolds for the good old American Life. And then we'll go ahead and create 11. Now we can add some integrations here like our Google Calendar, Hacker News, Linear Notion, Perplexity, and Slack. I went ahead and added my Google Calendar. So if I click Start call hello there.
Feeling productive? Let me know how I can help.
What do I have on my calendar for next week?
For the week of July 6th through July 12th, it looks like you have Adobe VIP Gen AI community call on Thursday, July 10th from 9am to 10am and news roundup also falls on Thursday, July 10th from 1pm to 2pm Is there anything else I can help you with regarding your schedule?
Now? I imagine they're going to roll out a lot more integrations over time so it can connect to even more services. But if you know what you're doing, you can actually connect custom MCP servers and have it connect to other services using their MCPs if they've got them something brand new that just rolled out but pretty fun to play with and over time will probably be more and more useful once you start using it on things like your phone and maybe if they add it into more devices that you might have around your house like your Alexa. All right, Another AI agent that I've been hearing a lot about lately is called Emergent. This is an agent that will actually go and write code for you. And unlike some of the other tools out there like Cursor or Windsurf, which is an IDE where you sort of create the code and then sort of go back and forth with it, this is more of an agent where you tell it what you want. It will create a task list of things to accomplish and then check each task off and then even go back and double check its work. So let's go ahead and test clone Spotify. We can see that it is setting up an environment and it's going to take about one to three minutes. We can now actually see it running. It said, I'm starting to run the task to create a Spotify clone. Running a web search with query Spotify Web Player. We can see the description that it came up with. Now let me get some images for the Spotify clone. It's now finding images or creating them, I'm not sure now. I went away, watched a YouTube video, came back and it says your Spotify clone is done. So if I scroll up. You can see it started this at 9:56pm if I scroll down, it was done at 10.02- so it took about six minutes to run this whole process. And we can see all of the steps that it went through to create this for me. And I can click on Preview here. And here's my Spotify clone. Let's even open it in a new tab to see what it looks like in full screen. It's got a search, it's got a your library, but it didn't go far enough for me to actually click into any of these yet. So there's no obviously real songs on here. And building this app here used 1.294 of my credits. And you can see up here, I started with 10 credits total. That's how much you get on the free plan. And I still have 8.71 to continue to further build out this app with Emergent, a pretty cool AI coding agent that you can use totally for free right now. Anthropics. CLAUDE also rolled out a feature called Artifacts, which allows you to write apps, use the apps directly inside of Claude, and then even share access to those apps with anybody. And there's even like a little marketplace kind of thing here where you can see some of the apps that have been developed. So let's try this Join Dots game here that somebody made. I could click on View Full Chat and I can see the entire conversation that was used to create this game. Over here, we could see the original prompt with all the details and all of the steps they went through with CLAUDE to produce this app. And we can even test the app directly inside of Claude. So let's go ahead and do Easy Difficulty. And it looks like it's basically Connect 4 that I'm playing with a computer. So I put my pink piece in, they put their yellow piece in. I'll put my pink piece here. They analyze the board and they block me. Over here, I drop another pink, they analyze it, and they do a good job of blocking me. Pretty cool little app. And you can use these anthropic artifacts for free right now. And you can even create your own. If I click on the Artifacts item in the left sidebar, I can click on new artifact. Let's do a productivity tool. And then it asks some questions about the productivity tool that I'm going to build. I'm just going to tell it, I want a tool that can track streaks on daily habits so I can chain days together. And let's see if it gets what I'm getting at. I'LL go ahead and answer its questions here with additional context. And I am purposely trying to give as little detail as possible so I can see how well it will perform form. Now it's shifted into coding mode and we can see that it created this split screen where it's now writing the code on the right. And after less than a minute, we've got our habit tracker. So Habit street tracker, let's add run as one of the habits and it put a bunch of giant boxes as the streak. So if I do it today, it fills it with green. If I do it the next day, it fills it with green. Not quite what I was looking for, but I mean, not too bad on a single prompt. I gave it a few more details that I wanted the boxes to be smaller and horizontal. And how my streak tracker looks like this a little bit closer. Not amazing, but not bad for three prompts on a free tool either. Oh yeah, and I can actually publish this now and make it available for other people to use the app that I made. No need to deal with hosting or setting it up on some other server somewhere else. It could just be hosted right here on Claude servers and I can share it with people. Now let's talk about the new agent in town, which is the Minimax agent from Haylou AI. Once again, this is a true agent where it will create a plan, it will work through its plan, it will go back and double check its work, even adding additional things to that plan until it realizes that it concluded what it set out to do. I'm actually gonna go ahead and test this agent with something that I've been looking for help with. I'm actually trying to buy a new truck and the truck I've been wanting to buy, I've been having a hard time finding near me. So I'm gonna see if it could do this research and find me this truck. I'm looking to buy an F150 lariat with a crew cab. I gave it about the type of engine I want, the colors I want necessary feature, and then I said, please help me find options within 600 miles of me. Sort your findings by closest to San Diego. Defer this from San Diego, since that's where I'll be coming from. So after about 30 minutes of running, yes, this is actually been running for quite a while. You can see all of the various things that it did to get me this listing here. In fact, I was watching it a little bit. It even had errors in the middle, recognized its own errors, and then said, let Me fix those errors, and then continued on with the process. And then when it was all said and done, it gave me this giant list of information, including giving me the search results in multiple file formats. Let's go ahead and click on the PDF format. And there we go. It found a bunch of trucks for me in my area and sorted it by distance from closest to me to furthest away from me. It even pulled in the phone numbers of the various dealerships, gave me a list of questions to ask the dealerships, and much more. I mean, this is. This is super valuable. It did take quite a long time. These agents that actually work like real agents are designed for you to give it a task and you literally walk away while it goes and figures out that task for you. And Minimax can do that. And you get a certain amount of free credits to just go try it. I didn't pay for this. Now we're getting into some just really interesting sort of fun tools that are out there. And this is one that I came across called Spawn over at Spawn S P a w n co. I don't know if you remember the site StumbleUpon where you just keep on clicking a button. It would just give you a random site to try. Do you just keep on stumbling around the Internet until you find stuff that was interesting? This site kind of reminds me of that. It's a whole bunch of AI generated games, and I think there's some apps in there, but it seems to be really focused on games. And I can just click Next Destination and it's going to throw a new game in front of me like this. Start Snake and I can start the game. And we have pretty much what looks like a Snake game where he collects a banana. All right, that came to do it for me. Next Destination throws a random game in front of me. This time it looks like it's a AI chess game. Here's one that instantly looks like a Minecraft kind of game, but way more messed up. And I could literally just click through and find little apps that have been made. Here's one where I could increase the trail length and it changes the design a little bit. Here's some sort of shooting games that I don't know how to make it work. Oh, there we go. But you can also use this site to create games as well. So if I come up to the plus button, I could create a new app by giving it a prompt or an image. I'll say make game. That's somehow a mix between Tetris and Snake. Apparently it's opening portals to new worlds. And now it looks like it's building several different potential games. Tetris pieces that move like a snake will be completed in about 16 minutes. Snake game played on a Tetris grid with stacking mechanics that'll be ready in 17 minutes. Navigate Snake through falling Tetris pieces. Snake that leads to Tetris blocks in its wake. It's apparently setting all of these up, just kind of running behind the scenes. And I should be able to come back and play these various games in, well, about 15 or 16 minutes. I just fast forwarded for you and our four apps are done here. All right, we've got Falling Snake. Okay, so we've got a Snake that has to play the Snake game while also dodging falling Tetris pieces. So I'm assuming if I hit a Tetris piece, I die. Yep. And then with this game, if I want to share it, I could copy and share the link to it and they'll even sell me a domain name to it if I want for reference. The Falling Snake one was the only one that actually worked. These other three all kind of had issues that made them unplayable. Along similar lines, we have Websim, which can be found@websim.com and this is almost like up a YouTube for apps where people use websim, create little apps and the site will just randomly recommend a bunch of different apps for you to play. There's things like Windows simulators, Grow a garden game, Geometry dash, face maker, throw things at Timmy, a Google logo maker. Just the most random stuff. World War three simulator. And just like the last site we were looking at, you can create your own apps. And over on the right side here, you can choose your model and prompt it to create whatever app you want. I think for me though, my favorite part about Websim is just exploring and getting lost in the various random apps that are in here. If we go to the top apps, we could see some of the Most popular ones. YouTuber Simulator. Find the fake variant generator. Just multi pendulum simulation. You can just click into any one of these and just have some fun. Like we can mess with gravity on it and see how it affects the thing. Change the speed, change the number of pendulums. This is wild. Here's one called Everyone versus Everyone. So choose my three fighters. Let's get Steve, SpongeBob and Mario. So here's my team. Let's start our battle. I'm playing against Caillou, Smiling book and Milo slides. Okay. Alright. So each character has weapons, so If I select Mario, apparently Mario's got a Glock. Mario. A diamond sword, an iron sword, a. On an RTX 5090. Alrighty then. Let's use the Glock, I guess. And we'll shoot Caillou. And spongebob just got shot with a Glock. All right, let's go ahead and pick Steve. And he can use an RTX 5090 on Caillou. And now Caillou's dead. Let's try the Corridor. I wonder if it's just like an infinitely scrolling corridor for no apparent reason. I can't like, move or change or anything. Like, it's just an infinitely going corridor. What would you expect? Will you go to an app called the Corridor? And let's try a Windows XP simulator. I was not expecting that sound. We've got Minesweeper. Does it actually work? Oh, my gosh, yeah. Minesweeper actually works inside of this Windows XP simulator. Nice. There's no way Minecraft actually plays. Oh, it's like a fake Minecraft. Okay. And clicking. Oh, we can click the. I mean, it's a fake Minecraft, but it's a pretty convincing Minecraft. Honestly. The little menu at the bottom looks off because it's just colors, but the blocks themselves look pretty good. Mario. So I'm now playing a Mario simulator inside of a Windows XP simulator, inside of a web browsing simulator called websim. Okay, this is pretty. I'm not gonna. Anyway, that's what I got for you today. 19 different, completely free to use AI tools that do some pretty crazy, mind blowing stuff. Now, I want to reiterate what I said at the beginning, though. I have no idea how long these will be free. Some of them are just in an alpha phase or a beta phase, and they're not charging while they're in that phase. Others have, like, free trials where the first month you can play with it for free and after 30 days then they want you to upgrade. But I think the biggest advice that I give people who ask me about AI and what they should do, I always tell them, get out there and play with some of the tools, experiment with them. Find out what you're capable of when you try coding. Figure out what you can create when you try these image generators. If you see a AI video meme circulating, see if you can recreate it yourself with one of these tools. Learn about the tools, learn about what they're capable of. And that's the best way to be informed about what's coming and what's available out there. There and also what sort of superpowers you can have using AI as tools in your toolbox. So I hope you found this video helpful. I hope you learned about some new tools. If there's some other free AI tools that you think other people should absolutely know about and try, make sure you let us know down in the comments. I'm always on the hunt for cool new tools, especially ones that you can use for free and I would love to hear about what I don't know about yet. Thanks so much for nerding out with me. I really really appreciate you tuning in and watching these silly videos of me with AI tools. I really really hope you're finding them helpful and I will see you in the next video. Bye bye. Thank you so much for nerding out with me today. If you like videos like this, make sure to give it a thumbs up and subscribe to this channel. I'll make sure more videos like this show up in your YouTube feed. And if you haven't already, check out Futurtools IO where I share all the coolest AI tools and all the latest AI news. And there's an awesome free newsletter. Thanks again. Really appreciate you. See you in the next one.
More from Matt

AI News: Dots, GPT-6.1 Sol, Sonnet 5.5, Gemini 4, and everything you need to know









