Everyday AI Podcast – An AI and ChatGPT Podcast - Top AI Tools and Releases of 2025

Episode Date: January 13, 2026

There's hundreds of AI tools released every day. Most are garbage. But these AI tools and model updates were the BANGERS that defined the year. So what made our top list? Tune in and find out. Top AI Tools and Releases of 2025 -- An Everyday AI Chat with Jordan WilsonNewsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Top AI Tools and Releases 2025 ListCriteria for AI Tool Selection 2025Honorable Mentions: AI Tools ExcludedCanva Visual Suite 2.0 Feature ReviewChatGPT Atlas Agentic Browser CapabilitiesChatGPT Deep Research Agent OverviewClaude Code Terminal Coding Agent ExplainedClaude Opus 4.5 Model BenchmarksCursor 2.0 Multi-Agent Code Editor UpdateDeepSeek v3 Open Source AI ModelsGenSpark Multi-Agent Workspace FeaturesGemini 3 Flash Model IntegrationGemini 3 Pro Multimodal Model ReviewGemini Canvas Mode Interactive WorkspaceGPT Image 1.5 Advanced GenerationGPT 5.2 Pro Expert Reasoning AbilitiesLovable AI Full Stack App BuilderManus Super Agent Workflow AutomationChatGPT Pulse Proactive AI Daily BriefingMicrosoft 365 Copilot Agent ModesNano Banana Pro Image Reasoning CapabilitiesNotebook LM Updates: Video & Slides FeaturesPerplexity Comet Agentic Browser AnalysisReplit Agent 3 Cloud App BuilderRunway Gen 4.5 Pro Video GenerationSora 2 OpenAI Video Model InnovationsSuno v5 AI Music Generator ImprovementsGoogle VO 3.1 Flagship Video Model ReviewS/A/B/C/D Tier Ranking of AI ToolsTool of the Year: Notebook LM StudioTimestamps:00:00 "Top AI Tools & Releases"10:16 "Atlas AI Tool Overview"15:21 "Claude Code's Growing Popularity"21:02 "DeepSeek v3: Open AI Model"24:36 "Gemini 3 Flash Dominates AI"30:10 "Canvas Mode & Code Rendering"36:11 "Lovable and Manus Overview"38:47 "Proactive AI and Personalization"46:25 "Perplexity and Comment Advantages"51:30 "Sora:Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

This is the Everyday AI Show, the Everyday Podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live in Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. There's quite literally hundreds of AI tools every year.

And aside from our newsletter, I don't really talk about AI tools a ton here on the show. That's because for the most part, I think they can be a distraction. So why am I then bringing you the top AI tools and releases of 2025? Well, it's the second half of that. it's releases because I think what this has become over the years is, well, you used to just have chat GPT. You used to just have Google Gemini. And now these companies actually have sometimes a dozen or more different releases or models or modes. So now I think that going over a top AI tool or release list of the year actually makes a ton of sense. And actually, hey, shout out to you

guys, our audience, because I wasn't even sure if I was going to put this show together and a bunch he reached out and said like, hey, Jordan, time's ticking. We're halfway through January and you haven't even put together your top AI tools and releases of 2025. So shout out to you guys. I work for you. So let's get into it and talk about the top AI tools and releases of 2025. All right. If you're new here, welcome. My name's Jordan. This is Everyday AI. It's a daily live stream podcast and free daily newsletter helping everyday business leaders like you and me make sense. these hundreds of AI tools cut through the clutter and just focus on those that are going to leverage the best and grow our companies in our careers. So it starts here with the unedited,

unscripted live stream podcast, but to be the smartest person in AI at your company, our website is your cheat code, your everyday AI.com. Go sign out for the free daily newsletter. We're going to be recapping today's show and a whole lot more. And the AI news will be in there as well. So let's get into this. These aren't the best. AI tools. All right. I'm by saying they're the top, I'm essentially saying that they're the most popular. And these are for everyday people. All right.

So obviously, if you're in software development, there's probably going to be some tools that are better or top tools that aren't on this list. But there are some, you know, coding and software development tools on this list. So keep that in mind. Or if you're, you know, a creative, right? You know, mid journey as an example, didn't make our list. But if you're creative, maybe mid journeys for you. So keep that in mind.

And also, these are only new tools, features, or modes that were released in 2025. So maybe there's a great, you know, AI tool out there that just wasn't really touched in 2025, hardly at all. Well, they're not even up for consideration, right? All right. And I've used almost all of them extensively. I think that there was only two on this list that I haven't used extensively. but I have used or tested them all. All right.

So there's actually a lot more. So if you want the entire list, which is like, I think ended up being like 48 to 52 somewhere in there. Just go ahead and repost this on LinkedIn. I'll send you the entire thing. I put it together in Gemini canvas, obviously. So it's interactive and it's fun. It's sortable.

It's a directory with some additional use cases, all that. So if you want the entire list, including those that did. make the cut, make sure to repost this show on LinkedIn. Speaking of those that didn't make the cut. All right, some of our honorable mention. So those that just barely missed it. All right. So 11 Labs conversational AI 2.0.

Google Flow, which this one hurt me, not put it on the list, as well as Google Pameli. Descript Underlord, Gamma 3.0 agents, which I love using base 44, Notion 3.0 agents. This one seems small, but the Gemini in Sheets equals AI function. This is so big. Literally a formula was so good. It almost made our top 25 list, all right? Claude code on web. And here's a shocker.

Chat GPT agent mode to not make the top 25 list. Shocker, right? So here's what actually made the list. All right. So right now, it's just in alphabetical order, but live stream audience, I want you to go ahead and get your voted. All right. And I'm going to do this probably again at the end. All right. So I'm going to go ahead and give a number. So you can just give the number for the one that you think is the tool of

year. But for our podcast audience, here we go. These are the 25 in alphabetical order. And then I'm going to explain probably spend about 30 seconds to a minute on each one. All right. So number one, Canva Visual Suite 2.0. Number two, chat gbt atlas number three chat gptee's deep research number four quad code number five claude clod opus four point five uh number six cursor two point oh number seven the deep seek v three uh b three two two in the r one family of models uh number eight gen spark uh number nine gemini three flash number ten jemini three pro number eleven gemini canvas mode number twelve gptt uh one point five or sorry GPD image 1.5.

Number 13, GPD 52 Pro. Number 14, lovable. Number 15, manis. Number 16. Oh, that's wrong on there. That's an old list. I'll have to give you the new number 16 here in a minute. Sorry.

Number 17, Microsoft 365. Co-pilot the new agent and office agent. 18, nanobanana pro. 19, the new notebook, L.M. studio updates, including Nanobanata Pro, 20, perplexities comet, 21, Replit Agent 3, 22, runway, gen 0.45, sorry, gen 4.5, 23, SORA 2, 24, Suno v5, and 25, V0.3.1. All right, and sorry, number 16 was chat GPT pulse. All right, so forgot about that.

That was the last second substitution and switch because originally I had MCP, but MCP was actually at the end of 2024. All right. So here we go. And as I go along, especially for our live stream audience, I'd love to hear what your vote is. Because at the very end, here's what I'm going to do. I'm going to rank these. So I'm going to go through and tell you them in alphabetical order.

But at the end, we're going to rank them in the old SABCD. All right. So I don't. I don't know if you've ever seen these charts before, right? They're pretty popular online. I obviously did a little vibe coding of this one in Google Gemini Canvas mode, but essentially S is your best. Then you have A, then you have B, then you have C, then you have D.

All right. So at the end, I'm going to rank all these accordingly. I'm going to try to put about the same number in each category. And this is, again, this is for me. But also, this is my advice for non-technical business. leaders as well, right? So keep that in mind as we go along. But live stream audience, I want to hear from you. So as I say each of these, let me know S is the best, a, B, C, or D. All right. So number one on the list, alphabetical order is Canva's visual

suite 2.0. So let me know, is that an S, A, B, C, or D. All right. So here's what the new visual suite 2.0 does. So unlike Canva's core content types, It kind of brings everything into a single visual suite workflow. So it focuses on scaling content production and it adds a build layer for interactive and functional content via natural language prompts. So essentially there's a bunch of these new features all over Canva. I use a lot of them every day. But I'm mainly just using some of the more simple, some of the simple ones,

right? The, you know, like magic text grabbing, which I really like by the way. So, you know, this is kind of like stacking, right? But I love using kind of GBT15 image or Gemini's Nanobanana Pro. And then when there's little things that aren't quite right, you know, they're not good at iterating. So bring it into Canva's new Visual Suite 2.0 with Magic Studio. And simple things like that are great for Canva. Everything else, if I'm being honest, right, the image models, the coding.

I don't know. I think that's just more if it's, that's all you have access to. But otherwise, I think there's better tools. But just because Canva has so many users, you know, I do think from a popularity and top users, you have to put it in. And this really kind of tries to push Canva beyond just a design program into more of a just multimedia creation suite. All right.

Number two, chat GPT Atlas. All right. So, and a live stream on, I see if it could, right? because I want to go through and see your guys's votes too, right? So if I just see a bunch of SABCs, I won't make sense. So try to put like chat GBT Atlas S, chat GPT Atlas C, right? Right, right, whatever you feel it deserves. All right.

So what the heck is Atlas? Well, this is Open AI's agentic browser. Um, did come out in 2025. It's based on chromium. Uh, and it's great for a lot of things. But, you know, probably the most unique thing is it can sync to your chat QVT data and vice versa, right? So chat chbtee can actually take advantage and understand and use the memory of your Atlas browsing and then vice versa.

When you're using the Atlas browser, you can take advantage of all your chat chitpity memory. So it's it's one of those things. There's a couple like weird UI, UX things that I wish were a little better inside Atlas. So if I'm being honest, what I use Atlas for a lot right now is when I'm scoping different use cases for every day. AI and I usually scope them across at the same time, comet from perplexity, which we'll talk about, Atlas, and then also the new Claude Code extension, which, oh man, that should have made my honorable mention list potentially.

So yeah, it's an absolutely great tool, but I think the real benefit of Atlas, obviously if you are a power chat GPT user, being able to use that and also agent mode, right? So agent mode didn't, make the list, but it technically makes it through chat GPT Atlas because that's a huge feature, obviously with an agentic browser, is having an agent that can navigate the web for you. So if you don't know, if you haven't used it or if you missed our show on it, right? So not only can it take

advantage of that chat GPT memory, but it can also just browse the web for you, both in the background and with an actual agent doing it. So, you know, clicking on interfaces. So, you know, as an example, we just launched our, you know, no one knows this. We just launched our community a couple of days ago, the inner circle community. I had the perplexities comet browser and the Atlas agent browser. I had them go through the entire workflow. They signed up for an email address. They went through. They signed up. They went through the onboarding flow. They clicked around, you know, here's what made sense. Here's what didn't write. Great like QA type of things that you can do with these agentic browsers. All right. And if you,

did miss the show. We went over this in episode 637 and we went hands on with opening eyes Atlas browser. All right. Next, Chad Gepti deep research. So this actually made the cut for early 2025. Just so you know, Google, Gemini's Deep Research was technically released in late 2024. Otherwise, it would have made the 2025 list. But ChatGPT's Deep Research is absolutely a gem. So if you haven't used deep research, you should start using it, FYI. It is really, really good. So essentially, it is an agentic browser or an autonomous research agent that goes through dozens of different sources and then delivers a comprehensive cited report.

It does take, depending on the query, depending on the sources, anywhere from five to 20 minutes, but it's absolutely worth it, right? If you're not running at least a couple deep researches a day for personalized or individualized learning, right? whether and don't take the word learning personally, right? You're like, oh, okay, Jordan, I'm not a student. I'm not an intern. Okay, well, you should still be learning constantly because the rules of the game are changing

by the day, right? Because AI is changing how business works. So we should all be learning on a daily basis. So I'm not kidding. If you're not running multiple deep research runs, whether that's in chat, GPT, Gemini, Claude, whatever. you are 100% 100% going to be behind, right? Yes, you need to know, you need to understand kind of the harnessing and how it's set up and,

you know, how these agentic models work. But that's why we have all the episodes that do that for you. But this really what this unlocks is just the fire and forget research, right? But personalized and individualize. So, you know, I love doing this before I go get a coffee or if I got to go do a quick task, you know, somewhere else in the house or in most cases, like even right. now as I'm on this live stream, I literally have two different agents working on something right now on a task going through our old podcast episodes to put together our starter series.

All right. So more on that here in a little bit. All right. So live stream audience, Chad GPD, deeper research. What do you give it? Also, go listen to episode 470 if you want a deeper dive on that because we did the deep dive on deep research for you. All right.

Number four, Claude Code. So, yes, I said this is for everyday people. And this Claude Code is probably the most technical. So we do have lovable on here, which I think is actually a little more beginner-friendly, but especially even with the new feature that was just released yesterday called Claude Co-Work, which is essentially a non-technical version of Claude code. All right. I couldn't put Claude Co-work on the,

the list because it just came out yesterday. All right, but this is what Claude Code is, and I think it's huge. So it is a terminal-based coding agent from Anthropic that understands entire codebases and executes development tasks autonomously. So it can read

the full code-based context. It can write your code, run commands, debugs, can manage Git pulls, your repos, all those things. And essentially, you can just build features, right? So obviously, if you're in software development, product management, et cetera, you probably already know Claude code because it's becoming a

de facto coding tool. But it, I think, right? And we'll see, right, my hot take as, you know, up until at least before yesterday, is that most non-technical people were going to be using Claude code. So maybe that won't be the case now with Claude Co-work. But I do think that most people will be using something like that. The trajectory of Claude code over the last couple of weeks has been something like I really haven't seen. In the last, you know, aside from the chat chbtee launch, it is going viral, which is strange because it's been out for many months.

But it's really picking up a lot of steam because I think people are realizing that you can use it for non-coding tasks, right? And it does really, really well at those. So this is best, though, for feature development, multi-file, refactoring, debugging, learning new code bases, et cetera. All right. Yeah, live stream minds keep like as we go along. Let me know. You know, is that an S, A, B, C, or D for you?

All right, number five, Clata Opus 4.5. So the combination of these two really good. So this is Anthropics' most intelligent model. It was the first to exceed the 80% on the suite bench verified, which is now the D-FAC, kind of software engineering benchmark. And it also has great computer use capabilities. And it operates graphical user interfaces like a human. Right.

So what this unlocks? Well, desktop automation. So we just talk about as an example, Claude, Claude, Claude, co-work. But, you know, you being able to do automation between your desktop, a web browser, terminal, virtual computer, right? And you're like, okay, Jordan, this sounds very technical. It's not.

Trust me, I do think probably most people this decade, that's going to be most people's jobs, right? What I just said there, you know, orchestrating agents, you know, that are using terminals, virtual computers, you know, repos, you know, interacting with graphical interfaces online for you, right? I don't think these are things humans are going to be doing, right, using the web, at least not like we do them today, right? Right now, humans use computers. We use internet. We use SaaS programs. I think agents are going to be doing a lot of those things, and we're going to be overseeing and orchestrating those agents. Anyways, Claude Opus 4.5 is probably one of the reasons why we're getting there.

So I talked about the meter metric last week on the show, and Opus 4.5 has been the model that has taken the big. biggest step from its competitors in terms of being able to work the longest autonomously and still doing a certain percentage of work at a human level. All right. And we did go over Opus 4.5 in episode 662. I will say this, though. I did that episode right after Opus 4.5 came out. We put it through some of our normal kind of, you know,

informal benchmarks. It did poorly, I will say this, but looking back at it now, I do think Anthropic has made some small under the hood updates to 4.5. Also, there's probably some access issues at the time and the rate limits stink on the base paid plan. I actually had to be grudgingly because of Claude code and now Claude co-work and the Claude extension. I did be. begrudgingly have to upgrade to the $100 a month max plate. All right. Number six, cursor 2.0. So the 2.0 update pretty big because it brings the composer and the multi-agents to the forefront.

So if you don't know a cursor, it's an AI native code editor rebuilt around agents, featuring cursor's own composer model. So Composer completes tasks in under 30 seconds, which is four times faster than the old version of cursor. And you can run up to eight agents. in parallel. You know what? There's so many, right? These are all like a lot of these are on the,

the top 50 list, but man, like, I'm trying to think, man, like so, it was so close, like codex, so close to missing, making the top, it was hard to make some of these cuts, right? The fact that opening eyes codex didn't make it, their agent mode didn't make it, Google's anti-gravity, you know, some of Google's other, you know, coding tools, even their, you know, simple coding tools like Opel,

There are so many good releases, especially from the big companies, right? Open AI, Anthropic, Google, and Microsoft, so many good releases. Anyways, back to Cursor 2.0. You know, what this unlocks, well, it's the ability for multiple agents to tackle this same problem and then the ability to pick the best result, right? So kind of like a mixture of models. Weird. My prediction from last year, not a mixture of experts here.

That's technically a mixture of models setup or kind of a hybrid between the two. So you can plan with one agent and then build with the other. So this is best for parallel AI development, large code-based refactoring, and kind of coding teams wanting concurrent agent workflows. Number seven, this is the DeepSeek V3-2 R1 family. So V3 came out technically in December, but V3-2 in the R1, which is Deepseek's reasoning model, came out in 2025. So if you don't know, these are check.

Chinese open source models that in a lot of benchmarks are at least top 10 and matching a lot of the frontier kind of AI models out of the U.S. at a dramatically lower cost. So some new things, well, that's R1's Deep Seek's thinking model. They're more transparent chain of thought shows reasoning steps, open weights for local deployment, and very cheap API prices on the dev side. So I will say this. Adobe just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the All In One Creative AI Studio.

Powered by Adobe's Creative Agent, Firefly AI Assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the Assistant. The Assistant orchestrates multi-step workflows, drawing on 60-plus programs, and grade tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom Express, and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks, like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible so you can refine, redirect, or take over at any time.

You stay in the driver's seat as the creative. creative director. Adobe Firefly AI assistant now in public beta. See it today at firefly. Adobe.com. I'm not using Deepseek. And I think for the most part, you probably shouldn't either, especially if you are a legitimate business in the U.S. I probably wouldn't be. Right. There's a reason why in certain states, Deep Seek is banned. In certain countries, Deep Seek is banned. Right. I won't get into that too much, right? But I will say, just because it made the list doesn't mean I'm recommending it. Again, when I say this is the top,

these are the most popular or most used, you know, features or modes from some of the big companies. And yeah, I'm not going to get into that, but just go ahead. I did the deep dive. I went through and did all the debunking for you.

But go listen to episode 460. The deep seek deep dive is an AI Sputnik moment or a national security threat. Yeah, probably more of the latter than the former, y'all. All right. Number eight,

Jen Spark. So if you don't know Jen Spark, They actually had a big partner announcement with Microsoft at Microsoft's conference, but it is a multi-agent workspace agent that distributes tasks across specialized agents working in parallel. I know I just said the word agent a lot there. It's much different than something like a cursor or an anti-gravity because these are essentially just you connect your data sources and it's a front-end agent, right? So, you know, this isn't using, you know, terminals and, you know, running code. Nope. You're just typing in a box like check.

ITBT, it connects to all your data. Then these agents run out and do different things for you and that connects to different models. So a lot of new things. So they have research agents, coding agents, analysis, creative agents, et cetera. But essentially, this gives you a team of AI specialists instead of a single agent or a single AI assistant. So you get multiple perspectives on complex problems. So, yeah, pretty good. Jen Spark.

All right. And yeah, live stream audience as we go, Jen Spark. What's a gig? What's your vote? All right. Number nine, Gemini 3 Flash. And this is pretty big for developers, but mainly Google's AI mode, right? So you might be wondering, like, what does Google's AI mode have to do with Gemini

3 Flash? Well, Gemini 3 Flash, number one, is a ridiculously good model. All right. On some third-party benchmarks, it's actually a top five model in the world, right? So if you don't know what it flashes, it's essentially the small version of Gemini 3 Pro. And usually these small or the mini versions, you know, there are many, many steps behind. Not the case for Gen.

I 3 Flash. It is a almost state of the art model, right? It is that close. But it's technically free to use because if you have a Google account, even a free Google account, you're just logged in, right? And you just Google something and you go to the AI mode. Well, now AI mode is so much better because of Gemini 3 Flash. Whereas before, it was running on Gemini 2.5 Flash, which again, not a terrible model, but a huge jump.

All right. So this is obviously great for developers. But the reason I'm including it on this is because this is probably going to be the most used, aside from some of the chat chivit tools, right? Because they have 800 million weekly active users. And for those chat chitbts that are at least you have a little bit on the free side are going to be used a little bit more than this. But Gemini 3 Flash worldwide is probably. going to be one of the most used models because AI mode is powered by it. And if you don't know AI mode, I tell people, always use AI mode.

Don't use AI overviews. AI overviews are not good. Use AI mode. It's one extra click. All right. This is best for everyday AI tasks, high volume applications, or anyone who found frontier models too expensive. So on the dev side, yes, it's obviously a great choice.

You know, a lot of developers use Gemini 3 Flash. It's fast, super fast, super cost effective. But even for non-technical front-end users, this is what you're probably going to be using when you're using Google AI mode, which I recommend you use all the time. Anytime my wife literally gets out her phone and starts typing anything, I'm always like, no AI overview. AI mode. Right. It's a rule in the house. All right.

Number 10, Gemini 3 Pro. So this is the big model. This is the behemoth, the newest, the biggest, the best. the baddest and technically the number one model in the world, both on LM Arena and on artificial analysis. So this is the best. So it did dethrone open AI, dethrone everyone. But it is multimodal by default.

So it can process video, audio images with equal depth, reasoning by default. On the API side, there's a million token context window, which is a ton. And if you are on one of Google's higher plans, you do get the deep think mode. I didn't include D. I could have included Deep Think as a top feature. But because it's only available right now, I believe on the ultra plan, I'm not going to do that. All right. So what does this unlock?

Well, a lot, right? This unlocks the future of, I think, AI models, right? There's a reason that Google has absolutely dominated 2025. And I think it's because, well, number one, their models are really good. All right. They had a very bad release with BARD. in the original Gemini, they were kind of caught tail between the legs in 2024. And they just head down, shipped, and crushed everyone in 2025.

And it wasn't really close. If you just look at progress in 2025, Google won. And I think in large part, the way they finish with Gemini 3 Pro, it is an extremely capable model. But like I said, the fact that it is multimodal by default, it understands the physics in video. Right. So not only can it create video. right it you know if you go to Gemini you can create video

Gemini 3 Pro uses Vio but it also understands video people don't know that right it can ingest video not just transcripts it can know what's going on like you know if I was sitting here you know if I throw mine in there it's gonna be you know it'll describe me it'll say oh I have a Nike shirt on whatever it understands everything but this is best for complex multimodal analysis video understanding log document processing, enterprise intelligence, et cetera. For me, I love using Gemini 3 Pro on AI Studio, another top 50, but not top night, quite top 25.

I love the new vibe coding updates from Google AI Studio and the team over there, which I talk to a lot, but sorry, y'all, this is too much good stuff. Couldn't crack the top 25 for Google AI Studio, but personally, I'm using Gemini 3 pro in AI studio. And speaking of that team over there in Gemini 3, if you didn't want to know, bit more. Go listen to episode 656. That was with Logan, Kill Patrick, and that was like an hour after Gemini 3 Pro was released. Had a very quick chat, but great chat with Logan. All right. Number 11, and let me know as we go along, y'all. This, I can't believe that

Gemini Canvas mode number 11 was released in 2025. It feels because I use it, all day every day, for whatever reason, it feels like it's been out for like two and a half years. But yeah, Jevonai Canvas mode was actually created or sorry, released in 2025. So this is essentially an interactive workspace within Gemini for creating and editing documents, code, and apps in real time. This is another one of those that was kind of closely aligned with the rollout of Open AIs canvas, which was released in 2020. for all right so jemini canvas was early 2025 so like open a i or chat tp t's canvas it does

kind of serve a dual purpose one is just live you know document editing with gemini so it's more of like you're working with a google doc and then you know the input which is i think in general this is great uh with chat gptt and google gemini right more people should be using canvas mode for exactly that so let's say i don't know you're having uh you know jemini or you know jemini or or chat chivity, you know, create a big, I don't know, outline for you, something like that. And there's, you know, a couple things in the middle wrong. What do most people do? Oh, go change, you know, paragraphs four and seven.

And then what do large language models sometimes do? Well, they'll change four, seven, and then they'll change the intro and they'll change the index and they'll change all these other things too, right? Well, use canvas mode, right? Because you can just highlight certain things. And essentially, it's co-editing. But the thing I use canvas mode for, which is great is, well, it can render, code, write and render code in real time. So everything from, you know, HTML, React prototypes,

web pages, infographics, quizzes, so many things. It is quite amazing. And the little tier sheet that I'm going to show to you guys here at the end to rank everything, obviously made with Google Gemini. All right. And if you want to know more about Gemini Canvas mode, episode 554. Number 12, here we go, live stream audience, let me know. What's your vote? This is GBT Image 1.5. This is OpenAI's latest dedicated image generation model with improved quality and speed over GBT Image 1.

And, well, what is this on lock? It's the best for, marketers, advertisers, communications professionals, anyone, right? And it is really good, right? Obviously, when it comes to kind of nameshare, I think that Nanobanamo, Pro, took a lot of the kind of media or social media attention when it came to AI visuals. But a lot of people don't know this.

On LM Arena, text to image, GBT Image 1.5 is actually better. And these are blind, you know, kind of like blind taste tests. You put in a text prompt, you get two images, and you vote for which one's better. So, you know, there's a lot of things that nanobanana pro is better at than gvt image 1.5 but it is extremely good right and if you remember the original um you know gbt image you know a lot of people uh thought it was not that good aside from you know there's some you know fun cutesy things that went viral but aside from that it wasn't very good uh gbt image 1.5 very very good all right so i'm going to put the the numbers here again on the screen for

the live stream audience as I take a sip of my water here. But go ahead. What is you only get one vote. What is your vote for tool or update of the year? And we're going to keep this going a little faster now because I don't want this to accidentally turn into an hour long show. All right. 13.

The model you go to when you just need it right. All right. That is GPT 52 Pro. All right. This is a. Again, thinking mode. Don't use, y'all, don't use the free version of GBT, chat GBT, GBT 52 instant, GBT52 auto.

Don't do it. GBT5 thinking is good. GBT 52 Pro is literally yesterday, solved some physics problem that humans couldn't solve. All right. This is good. All right. This is when you need something right. So this is OpenAI's flagship.

model with extended reasoning capabilities. So in the family of models, there is kind of three different tiers. There's the instant, the thinking, and the pro. But don't use the auto, the instant. You should be using thinking. But if you have the time, right, use pro. GVT5 Pro, it can be very slow. And I'll tell you this.

My usage of chat, GBT's deep research has actually gone down because GBT5,2, Pro, I think, is just as good. But then you can kind of tap into some other features by using this model because you can, as an example, use canvas mode in there. You can use a, a GBT can be powered by GBT 5.2 Pro. So there's so many benefits to maybe using GPD5 Pro where you might usually use deep research. Depending on what plan you are, you know, different limits.

I get it. It's a little bit of give and take. But I mean, this is human expert level performance on complex. complex task, but only when you are using the thinking mode, obviously. But it's best for complex reasoning, financial modeling, multi-step projects, professional codes, coding, et cetera. Just obviously always make sure you're using the GP-T-5-2 thinking or GPT-5-2 pro.

Do not use the default version. I've talked about this many times on the show. It is the 24th best model in the world. You wouldn't want to do that, right? If you have the choice, would you want to use the, 24th best, you know, dentist in your city or the first, right? Probably the first. All right.

Number 14, lovable. All right. So this, a lot of good updates in 2025. But if you don't know, Loveable is a platform that builds full stack web applications from just natural language. So they had a lot of improvements in 2025, the ability, I believe they had the cloud, the back end for off everything, right? Whereas before you kind of had to connect it to.

you know, you had to duct tape kind of whatever you created in lovable to, you know, some different back end, so different off services. So now they kind of have everything with the lovable cloud where it's just kind of ready to go. And they also now have the agent mode at the default. So it's autonomous multi-step building and it generates front-end, back-end database deploys and enables conversational iteration. So essentially describe an app, get it deployed. And this is best, I think, for non-technical founders, you know, cursor. And, lovable two very different things, right? Cursor, I think, is more for software engineering teams, very serious developers. I think lovable can still be used by those people, but lovable, I'd say, is probably for

maybe if you're a non-technical person, but you've had an idea or an app or you run a company and, you know, I don't know, maybe you're paying $50,000 for, you know, a piece of software each year and you barely use it. And you're like, maybe we could just build something like this on your own. And, you know, you try to get something together, get an MVP together in an afternoon. Loveable might be the thing to look at there. All right. Number 15, Manus, just the general purpose super agent.

So it is the first kind of big name general purpose autonomous agent that can just deliver complete work products from gold descriptions. So this was launched in 2025, but you can literally have, they have the wide agent feature where you can go have like a hundred agents researching at the same time they all spin up their own virtual computer to execute and handle complex multi-step tasks independently and it's blown up right so they were actually just acquired by meta right not even a year after their you know public launch so they did hit reportedly a hundred million dollars annual recurring revenue in eight months which is an astronomical amount of revenue i use manis i'm on their paid plan it's really good uh What I need to personally get better at menace for is sometimes I just hand it off everything

when I probably shouldn't, mainly because of the credits, right? So if you also, like I have subscriptions to obviously everything. Manus is great, I think, at more quickly logging into all your services, right? That's what I am going to start using it more for. But, you know, on the front end, on the research end, unless you're trying to get, you know, lead information, which it's really good at, by the way. but don't waste your credits on, you know, typical research, you know, go do that in like OpenAI, deep research, you know, Claude, Google, et cetera.

But it is good for data analysis, resume screening, travel planning, any complex tests that you would designate to an assistant. If you wish you had an assistant, Manus is pretty good for that. And we did go over Manus and Jen Spark and a lot more. In this episode, I think, was actually kind of underrated. So episode 613, which is AI agents from automation to super agents, 10 AI agents you should know in 2025.

So go listen to episode 613. All right, number 16, the correct number 16, chat GPT polls. So I did still include this, even though it's currently only available to those on the $200 a month plan. I did so because I think it signifies kind of the next step of AI, which is proactive AI that just does task for you, even if you don't ask for it, and just delivers you what it thinks you need.

So that's essentially what chat GPD pulses. It's a proactive daily briefing where chat GPT researches overnight and delivers you a personalized morning update. I don't like it that much, if I'm being honest, but I do think just the number of users that Open AI has, and I'm a little bit different. I say I'm a very unique large language model user.

I'll say that. A lot of times I'm really, trying to test things and break things. And so I was getting a lot of reports that were not maybe relevant. It was maybe relevant to the usage, but not relevant to what I wanted. So you can obviously provide feedback to chat. ChbT pulse, right? It can go through your, your email, your calendar and, you know, tell you what's important.

And, you know, oh, you have a meeting with this person and, you know, does some research for you. I don't know. To me, even the more feedback I gave it, I wasn't getting something great. And probably the reason is, is because I am really good at chat GPT. And a lot of people don't know. here's a secret. Lean in for the secret. Agent mode, you can schedule agents, right? So at least for me, I was able to get much better, I was able to get much better results, especially when we still

had connectors. We have apps now. They work a little different. But, you know, the agent connecting to my data on a scheduled run was way better for me than a, the Chad ChbPt Pulse, because I just had more control over it and I had less control over pulse. So, you know, I think for people that are more casual users, or, you know, even people who are, you know, power users, I would say I'm just trying to break things all the time and reverse engineer things. But for everyone else, I think it's great. Chat Chitpity pulses.

All right. Number 17. And let me know live stream audience. Keep getting your votes in. Number 17 was the Microsoft 365 co-pilot agent modes. So this is an AI assistant. embedded throughout Microsoft 365 with autonomous agent capabilities. So the agent mode completes multi-step tasks across office apps autonomously,

including the memory work IQ, that it maintains persistent context and learns performance. And there's also the ability to access this in co-pilot studio. So this really gives you the ability to delegate complex workflows across Word, Excel, PowerPoint, Outlook, and Teams. And this is obviously best for Microsoft Native Enterprises. And we did go over this in episode 621, hands on. A lot of these, by the way, these episodes that I'm dropping, probably best if you watch them, because a lot of them are hands-on demos, not all of them, but just FYI.

If you ever want the video version, you can always get that for free on our website at your everyday AI.com. All right. Number 18, we're almost at the end of the list here. Nano Banana Pro. I already kind of referenced this when we were talking about GBT for image. 1.5. But nanobanana came out. Amazing, right? I think nanobanana was the first time that we looked at images and we're like, wait, even as a former photographer, right? I have four professional

DSLRs, you know, in, in my closet. And I've taken more than a million photos over the years. Nanobanana, I could tell the difference. I could say like, okay, yeah, this is AI or nanobanana. The average person probably couldn't. Nanobaneda Pro, no one can't. No one. Not if you're an AI expert, not if you're a photographer. No one can tell the difference anymore. Nanobanana Pro and GPD15 image for certain use cases are that good.

But Nanobanata Pro was the first model that got to that level where it's like no one, no one can tell anymore. Even if you're trying, you can't tell. So this is best in class text rendering. finally solved, right? Took a while, but, you know, now you can finally get that. You know, it reasons, which is great. This is a reasoning model.

It's part of Gemini 3. We went into this more in depth when we went over Gemini 3, uh, pro, but it can reason about an image, which is completely different than, you know, a diffusion model that doesn't think it just, you know, kind of spits out and fills in pixels. This understands what's happening in the photo, what should happen in the photo, right? It's extremely accurate. You can also use it. You can upload 14 different references and it can combine different things.

So I don't know if you had a bunch of products, right? Maybe you work in an e-commerce company or something like that. You're like, oh, man, we didn't get any of these into shoot, but we have individual products. Well, you can literally just combine things and then combine different, you know, products with different, you know, humans or models. And then, you know, fill it in with text. It is that good. This is obviously best for people, you know, trying to do marketing assets with. text infographics it's absolutely bonkers at localized visuals product mockups etc number 19

this one notebook lm updates including nanobanana pro in video overviews and slides so notebook lm actually had a ton of updates this year that weren't related to nanobanana they redid their studio completely So now there's personalized and customized reports that you can save templates for, which are really, really good. You can create unlimited versions of those, the AI podcast, right, the deep dive overviews. There's different versions of those. Again, you can create however many you want because you used to only be able to create one, right? It's such a huge unlock, even just those things alone, just the studio.

But then you throw in Nanobanana Pro, right? So now, NoBook LM has video overviews and those visuals are powered by Nanobanana Pro. And then they also have slides that are powered by Nanobanata Pro. Yeah. When I say like, I don't want to be the product manager PowerPoint. Yeah, don't, don't want to be. Right. Not good because Nanobanato Pro is so, so good, especially combined with notebook LM and being able to run slides.

in there. It is bonkers. Right. So I'm looking at this one live stream on it's, you know, who's going to give, uh, you know, I'm wondering if anyone's going to give notebook LM, uh, number 19 here at D. We'll see. Now, we did go over this in episode 652. That's another underrated episode. Go listen to that one or watch that one on our website. All right. Number 20, perplexity comment. So this is an agentic browser from perplexity, uh, like Chad Chubbiz Atlas. It is chromium based. So like chat chad chvety atlas, you can, you know, sign in with your Chrome profile. It'll sync everything over, your bookmarks, your passwords, your extensions.

Same thing with perplexity's comment. If I'm telling me the truth, there's not a ton of blanket statements or blanket differentiators I can make between perplexity and Chad ChbT Atlas. The thing I can say, the big benefit with Chachypte Atlas is it connects to your data. and I think it's a little better on the agent mode. Perplexity for things that you don't necessarily need a graphical interface for, you know, if you don't need an actual agent clicking, you know, making 10 clicks on a page for everything else.

I think perplexity comet is a little faster and does a little bit better at long range tasks in my testing. Like I said, I'm constantly testing Chad, ChpT Atlas, Perplexity Comet, and then just Chrome with the Claude extension. And perplexity comet is usually the best for most use cases and the fastest, and it does better over long range tasks. So this is great for research across many sites, price comparisons, data collection,

you know, any task that would normally require multiple browser tabs or it require you to go to a bunch of websites back to back to back and find all this information. Perplexity comment, great. And we did go over five. business use cases for comments in episode 614. So go check that one out. All right. Probably the one I've used least on the list, but I am bullish on it regardless, because

it is really good. Replit agent three. So this is a browser-based AI agent that builds, deploys, and hosts applications entirely in the cloud. So now with natural language, you can go straight to a deployed app, describe what you want. Agent builds and ships it. And with integrated hosting, you have no local setup required.

So I would say that this is, I wouldn't say the more technical version of lovable, right? And Replit's been around for forever, right? I would say Replit is the, you know, kind of the AI clone of Replit. Not in a bad way, right? It's just non-technical, right? You can get more technical in Replit if you need to. But with Agent 3, you don't have to if you don't want to, right? They actually had a funny little ad a couple weeks ago that we shared in the newsletter with Shaq, right?

Just literally using his voice and creating apps, you know, full front end, full back end. But what this unlocks is essentially being able to go zero to deployed in the browser with complete dev control that's accessible anywhere. So great for, you know, whether you want to learn to code, deploying simple apps fast, rapid prototyping, you know, developing without a local environment, whatever it is. Replit agent three really, really good. All right. We got a video model. There we go. Runway Gen 4, 5.

So runway was technically one of the OG video models, right? Before we had a VO, before we had a SORA, we had runway. So this is actually great for professional grade AI video generation for, you know, both filmmakers and creatives. So what's new this year? Well, Act 1, which is impressive. So you can capture facial and body motion from your webcam and transfer that to AI characters. It's bonkers, right?

Such a cool feature. So if you are trying to do something, a little higher level, right? You know, trying to get the right facial expression on something. Maybe you're just shooting a little ad for your company, right? And you can't get your avatar or whatever you're working with. You can literally act it out, the facial expression, the body expression, and it will do that. They also did add the multi-motion brush for selected animation, which is good. They've had this single animation brush for a very long time, so that's good to get the multi-one.

And then, well, what does this unlock? Well, controlled AI performances using your expressions, stable characters across frames, professional quality output, and obviously the new updated, you know, going from Gen 4 to Gen 4.5, pretty big boost. So who's this best for? Marketers, creators, advertisers, filmmakers, filmmakers. commercial production, music videos. Essentially, if you're a creative professional,

you should probably be using runway gen 4 or 5. Or at least trying it. All right, next on the list, SORA 2. So what is this? Well, this is Open AI's video generation model with native synchronized audio. That's the, you know, that's really what was unlocked in 2025.

You got to tip the cap to VEO, which we're going to get to here in a second, which was the first, right? But SORA, good. It's really good, right? The video model is very impressive. And it actually launched via an app, right? So yes, it's part social app slash brain rot, but still really impressive video model.

So you do have to sign up for it on via the mobile app, which is technically like a social network, but you only have to do it once and then you can just do it on your computer from there. So now you can do up to 25 second clips with native audio, including dialogue, to lips, sound effects, and music. There's a storyboard mode, which they've had since the original SORA, which again, a really underrated feature of SORA, best used, I think, on the desktop, obviously. They have now the character cameos, which are just called characters now, right? So you can upload yourself, and then you can then just have yourself do anything.

Or you can make your character public, or you can share your character with friends or coworkers. and they can use your character to do or say anything. Right. So there's obviously a lot of, you know, scary things that can be done. But if you're putting your character out there for anyone to use, well, you can maybe expect people to use it for bad purposes.

But I think this unlocks just so many creative use cases. I'm not personally using this. Obviously, I used it to test it a little bit. But I think this is best for short form content, social video, marketing clips, all those things. You know, so if your, uh,

if your company is trying to, you know, make it on TikTok, uh, you know, with vertical video, SORA is a fantastic tool for that. Me,

I could care less about social media, vertical videos. I don't scroll, uh, you know, vertical videos. So not my thing personally, but extremely impressive.

All right. And we did go over that Sora 2, uh, AI TikTok brain rot or your company's secret creative weapon, uh, in episode 623. All right. Two more. Number 24, Suno B5, the AI music generator. Essentially, without going into too much detail, there's a lot of new things,

but the major thing is quality jump, more natural vocals, better instrument separation. You can actually separate different things. Almost think of a song now that's layered and you can control each layer with text prompts. You can also, you know, upload samples of yourself like humming a tune, things like that, but Suno v5, again, this is one of those things where it's probably better than, you know, 90, I'm going to get some hate mail from this. Sorry, it's better than 99% of musicians, right? Like, essentially, if you're not, you know, on tour or on the billboard charts or, you know, probably Suno v5 is just as good or better if you're not one of those things.

right the music it creates bangers right bangers people didn't know this there was actually an artist that was breaking onto all the billboard charts and they found out later it was someone just using suno but i think that's going to become maybe kind of the norm because the music is that freaking good uh and this is an older one but it was just a fun interview uh that i had with the CEO of Suno a couple years ago, episode 207. It's probably a cringy interview because it's so old, but it was a fun one. I'll leave it at that. Go listen to it.

All right. And then last, but certainly not least, we have V-O-3-1. So this is Google's flagship video model, and it was the first to generate synchronized audio natively and really good. So what is new in 2025? Well, the native audio generation, including ambient noise, music, dialogue, sound effects alongside video, the physics understanding through the roof because it's Google in up to 4K resolution.

So this is true end-to-end video generation without the need for extensive post-production audio work, right? The ability for, you know, beginning frame, end-frame generations. Bonkers. So good. I wish I had more time of the day. I would be using V-O-3-1. I would be using Google Flow more.

But this is best for video content creation, marketing. essentially anyone wanting complete videos without editing audio or without needing a huge team. That took way longer than I thought, y'all. I am physically tired after this one. So many AI tools. But go ahead, get your votes in one last time. So live stream audience, I'm going to give you a second. Go ahead, vote for your favorite number.

Just put the number if you want, you know, you can, I guess you can click pause on a live stream too, right? Yeah, you can click pause on a live stream and then just finish, you know, you'll be five seconds behind. You're like a delay. Like that time, I had a Super Bowl party with two TVs and one was delayed and, you know, everyone's screaming over in this corner and everyone in this corner is confused, right? There you go. All right. Now, if you want the top 50, everyone that didn't make the list, make sure you repost this. All right.

So now I need to quickly rank these, even though we went very long. Let's go ahead and rank them. And like I said, with our little vibe-coded ranking system here from Google Gemini in the canvas mode. All right. Let's do it.

Let's get down to business. Where would you put everything, y'all? Where would you put everything? It's just tough decision. You know what? I kid you not. Guys, I never thought like people listen. I hear from people all the time like at Open AI, Google, you know, Microsoft other labs.

And they're like, oh, I saw you, you know, talked so great about this. And I was kind of saltier, you know, hurt. So it's like, I don't want to hurt anyone's feelings. But yeah, I'm just going to tell you guys the truth here. And just because something's a D, right? That doesn't mean it's bad. That means it still beat out hundreds of other. and by hundreds, I mean thousands of other extremely capable AI tools.

And again, remember one last reminder on this. This is for non-technical people, right? This is for your everyday business leader. Like I said, there's so many other great, right? If you're talking about certain niches, certain verticals, there's so many great tools. All right. Let's go quick here. This is a long one.

All right. Here we go. Let's start with, let's start in the middle. So I think lovable, great. I'm going to put that at sea level. Also on sea level, manis. I've been using manas a ton more lately. They've been rolling out some great updates in 2026 already that I'm loving.

Like I said, I got to get a little bit better at, you know, testing out the 1-6 light, the normal 1-6 and the 1-6 max. Right. Sometimes I just go straight max and eat through my credits. One prompt, A-through my entire month's credits. All right. Let's keep this thing going. You know what? Quad code.

Where do you think? Oh, where do you think Claude code is going to go? Bangor. S. Claude code. It is so good. All right. Canva, Visual Studio or Visual Suite 2.0.

That's D tier. Still made the list. Still use it technically. Every single day I do. All right. Let's go here. Let's go here. All right.

We're going to go Suno. On the D. Nothing. Suno, it's literally fantastic. I'm just thinking for overall use cases. I still think a lot of people are using music, but it's the best one by far, which is why I made the list, and it's extremely popular. All right, let's keep going.

Let's do chat GPD pulse. That's going to fall on the D. Again, I think if more people had access to it, if it was a little better receiving feedback, maybe a little bit higher. All right. Let's go so many options, so many options. All right, let's go Jen Spark. That's going to go D.

That was tough for me. Probably between D and C, but probably, probably right there. All right, let's go. By the way, Gemini Canvas crushed this thing. It's so smooth. There was other ones online, but they were all buggy. And I'm like, I'm just going to make my own, my little tier sheet, and I can rearrange them. All right, Gemini 3 Flash.

There's a lot of things I got to get to. All right. That's going to be B. That's tough. It is so good, so fast. It's going to be used by billions, literally billions of people. But I got to put it there. All right, Gemini 52 Pro. Gosh, it's got to go S. It's got to go top, right? It is the best model. Even though Gemini 3 Pro has become a little bit more of my workhorse, my day to day, when I need something done 100% and I need that chain of thought. I need that transparency. I'm going to Gemini 52 Pro and I'm not looking back. All right, runway, Gen 4, 5. We're going to drop that down D level.

All right. We're going to go chat, GPT, deep research. Ooh, this is tough. I'm going to have to go to A. Can't put everything on us. That's the tough part. All right. Let's go to Replit.

We're going to put that at C. Replit, Agent 3. So good. Deep Seek, same thing. We're going to put that at C. More than anything, it's not even necessarily the models, It's what it's done for open source, right? That's why I'm saying it's a top AI tool and release because it is impactful.

Not because I think people should be using it, but because it pushed all the other companies, right? We got an extremely capable GPTOSS from open AI, maybe not in response, but I don't know. They had to be influenced somehow by that. All right. Let's do. Let's do perplexity comment. This one's tough. This one's tough.

All right. I'm going to have to do for a perplexity comment. I'm going to go B level. I'm going to go B level with that. All right. And because it's so those, I mean, those two are neck and neck. I'm going to also put chat Chbety Atlas, B level. Let's start filling in the A level.

Okay. Oh, V-O3-1, man. The fact that that didn't make the S level, it's going to, it's going to A. All right. All right. Also going A, we're going to go to GPT Image 15. Where are we put in Microsoft copilot agent modes? I think that's going to go B level.

That's going to go B. All right, Sora, Sora 2. Okay, we're going to do Sora 2. I think maybe, we're going B or C, y'all. I'm going to go B for now. We might change that. I wanted to try to make these somewhat even, so I might change that. Oh gosh, Gemini Canvas mode.

That's got to go A. It's so good. Can I put it as? Probably not. I'm looking. I got smasher bangers only left. All right. Cursor.

I think cursor we can go C. Extremely capable. Still really good. Oh my gosh. This is hard. Okay, so I have five at D. All right. So I have Suno v5, chat Chb-T Pulse, Runway, Gen 4-5, Canva Visual Suite 2, and Gen Spark on D.

On C, I have Replit, Agent 3, Cursor 2, Deep Seek family, Manis and Lovable. And then on B, I have Gemini 3 Flash, Perplexity Comet, Chatt-GPT Atlas, Sora 2, and the Microsoft 365, co-pilot agent modes. And I have a couple left that I got to fill in. I mean, can all of these go S? Let me see. Okay. That would just be, I can do this, right?

I mean, Nanobanato Pro, got to go S. Claude Opus 4-5. Oof. I would catch heat for putting this aid level. It's a really good model. All right, it's got to go S. Gemini 3 Pro, got to go S. And y'all, for the second year in a row,

notebook LM, studio updates. Not only go in S mode, but tool of the year. Yeah. Shout out notebook LM team. Tool of the year again, it has to be. So last year, it was just the tool of the year because it was brand new in 2024. But the fact that in notebook LM now, again, it's grounded. All right.

So what that means is it's not pulling random information from its training data on the web. It's only what you give it, which is why, if certain companies, individuals are like, oh, we're not sure about large language models. Use notebook LM and you'll be like, oh, my gosh, this thing is, I'm never going to stop using this, right? They came out with a great mobile app. You can upload screenshots in the mobile app.

All this studio advancements that they made the ability to create unlimited deep dive podcast, the different versions of deep dives, the customizable studio reports. My gosh, the videos powered by Nanobanana in the slides, slides powered by nanobanana. My gosh, I mean, it is almost as good as Gemini, as chat chivity, as clawed, right? What you can create inside notebook LM because of these 20, 25 updates, it is on the same, right? If this was by another company, right?

I think maybe it doesn't get the love that it might deserve notebook LM because it's a Google product. And they're like, oh, well, you have, you know, Google Gemini. Well, okay. I mean, you have notebook L.M. That's powered by Gemini 3. You know, you have the videos, the nanobanana, the slides. It is that good. It is our tool of the year again this year. But in our newsletter, we're going to announce the people's choice, the audience choice. So I'm going to go through and I'm going to count whichever one has the most votes from the audience. And we're going to give that the audience award. All right. That was a ton of. to go over. I'm sorry, this turned into an hour long episode, but I don't talk about AI tools a ton.

So, you know, I'm just shoving it all into one episode a year. So that's why it's a little longer. All right. I hope this one was helpful. Like I said, make sure to tune in. What is? Yeah, tomorrow. We're going to be starting our start here series. All right. People have been asking for this for a literal since 2023. you know, they're like, Jordan, you have hundreds of podcasts. Where do I start? And I always have to like answer that individually. Well, I'm like, it's January. Everyone's trying to, you know, double down and learn about AI more.

So we're going to have a start here series. We're going to probably release two or three a week, you know, off and on. We'll see. But, you know, we're going to release them intermittently through January and February as we get done producing these. So these probably won't be live. They'll be a little faster. All right.

But these are great. Whether you're a beginner, you're starting at square zero, whether you're like, oh, I've been tinking around with a toy or two, you know, Chad JVT here, Gemini there, right? Or even if you're someone that's using it every single day, it's always changing, right? People are always like, you know, the analogy is like, oh, you know, if you were a history major, you know,

what changes in history? Well, not a lot, right? So even if you're using AI every day, I guarantee you, you are going to benefit it so much from our start here series. All right. And this is something we'll probably keep updating a little bit in the future as things change,

but this is our number one request. People are like, where do I start? So this is covering, think of this as like a 101 course in, you know, university. You got to get all your gen eds right before you take your special classes because, you know, otherwise those classes for your major aren't going to make sense

if you don't have the baseline. if you don't have your core classes. So this is the core class. If you want to graduate from everyday AI university, if you want to be an AI leader, if learning generative AI, large language models, was on your to-do list, was on your kind of New Year's resolution.

The start here series is for you. So make sure to tune in tomorrow as we kick that off. And make sure to go to your EverydayaI.com. Sign up for the free daily newsletter. Thanks for tuning in. We'll see you tomorrow and every day for more Everyday AI. y'all. Meet Firefly AI Assistant. Now live in Adobe Firefly,

the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the

outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adobie.com. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.