Google Flow is Google's AI creative studio for generating and editing images while using AI. It combines three Google models under one roof. First, you've got Veo, Google's video generation model, which is currently Veo 3.1. This is the engine that creates your video clips from text prompts. Then you've got Nano Banana 2, which is Google's image generation model evolved from Imagen, and this creates high-fidelity images that you can use as standalone outputs or ingredients for your videos. And then of course, you've got Gemini working behind the scenes to make prompting more intuitive, helping Flow understand natural language descriptions. And so this is the platform that everyone's going to use after Whisk has been discontinued. Firstly, let's talk about the pricing because if you're going to be using this, you need to understand exactly how much it costs. If you simply have a free Google account, you essentially get access to 50 free daily credits. So this can be used in Google Flow. So that does mean that in order to test out this model, you won't need to exactly have a subscription, but 50 credits won't actually get you that much because every single model does cost a certain amount. The one that is actually the most cost-effective if you want to

use Google Flow is going to be Google AI Pro. This is the tier that is actually at a reasonable price, around $20 a month, and you get around 1,000 monthly credits every single month that can be used inside of Flow. And of course, you do have the Google AI Ultra plan, which gives you 25,000 monthly AI credits to be used in Flow. Now let me show you guys exactly the credit cost per generation. Sometimes Google can and will change the pricing of certain models depending on depreciation and newer models as they come in and come out. So if you want to know how much something actually costs, click this where it says video or image, it might say either one, and then in this box right here, you'll be able to see exactly how much credits it costs. Now, these credits are essentially based on a multitude of factors. Either images or video cost different amounts of credits, and the amount of videos that or images that you generate will cost different amounts. So for example, let's say I switch to image, you can see on this tier, generating will use zero credits, but for videos, if I choose to generate three of them on the Veo 3.1 fast, that

works out to 30 credits. If I go down to one, that of course costs me 10 credits. It's important to understand that you double-check this before generating because there have been numerous times that I haven't realized how much it has cost me and I've ended up spending more credits than I would have realized. You can also change the model here to Veo 3.1 fast, and depending on your tier, it may cost zero credits. Of course, Veo 3.1 quality, that one is going to cost 100 credits per video. If you want to know exactly how much credits are going to cost you, it essentially gets cheaper the more you pay. So if you're on the Google AI Ultra tab, essentially, you do get more bang for your buck, but it is of course more expensive at around $200 to 250 pounds per month. I would say only get this if you're going to be generating multiple videos and clips every day. Otherwise, just use the standard Google AI Pro tier. Now something that is rather strange, free users actually get more credits than Pro in some cases. For example, when you figure out that free users actually get 50 credits a day, which actually does equate to 1,500 a month versus the Pro tabs' 1,000 monthly credits, it might seem like that is a bad deal, but do

understand the free credits don't roll over day to day. So that means that free users aren't able to generate certain kinds of videos, and they aren't able to stack credits for a good session when they need to burn through a certain project. So do remember that if you upgrade to paid, your free credits are immediately forfeited, replaced by the new monthly allocation. Remember, the credits don't roll over month to month for any tier. You either use them or lose them. If you're wondering about what your current balance is going to be, if you click the top right, you'll see your balance just underneath your name. And let's say for example, you've decided to run out of credits. If you want to get more, click get more credits, and then this is where you come to the section where you can buy more credits. And do bear in mind that if you do buy top-up credits, those are valid for 12 months from the purchase date for the Pro and Ultra plan only. If you are wondering, "Okay, I have credits, some are free, some of this, some of that, will I ever get throttled?" You will get throttled on Nano Banana 2 images after a rate about 100 a day. So do bear that in mind before generating multiple images. So if you've never used Flow before, essentially what you'll see when you come on here is probably nothing. But if you have used this before, what

you'll see is this grid. Now this grid is essentially every other project that you've worked on before. And here, let's say for example, yours is messy, you can quickly rename these in order to organize the chaos. So for example here, I can say this is going to be Google's AGI because I know exactly which video this is. And so you can actually do that. But by default, when you do create a project, it just stamps the date and time. So let's start a new project and I'll show you guys exactly how I use Flow. So once inside of a project, what you might want to do is of course either generate media or drag and drop media. If you want to add some media, click the top right button. Here, you can either upload an image or create a collection. Of course, at the bottom, this is where you essentially you have all of your prompts for images and videos. If you click this, you can see on the left you've got images. Here, you can select the aspect ratio as well as the amount of generations that you do want. And of course, the model that you do want. Nano Banana 2 is currently the best model, but of course I'm guessing once you're throttled, you'll probably switch back down to Nano Banana Pro or Imagen 4. Once again, you can go to videos. Here, it's a little bit more difficult, but I'll show you exactly how we can do that. Let's start with images first. Here, I'm actually going to upload an

image. Don't worry, images don't upload in a second, they do take a few seconds to upload, so don't think there are any errors. You can see I've got this image of a McLaren right here. Of course, as well, now that we have that image right here, I can actually start adding this to my box if I want to actually get started. So now, once you have uploaded your image, there are several different options. Here, you hit these three buttons, you click more, you can see you've got animate. This essentially allows you to add this and it will immediately turn the prompt box into a video prompt. So for example, if we want to turn this immediately into a video, we click animate, and you can see immediately we get this prompt here. Don't forget to always adjust these and quickly check how many videos that you'll generate. It will always pop up with two frames, one for the start frame and one for the end frame. But before we generate a video, if you do want to, you can integrate with these images even more. For example, what I can do is I can simply just add this to my prompt. And then what I can do here is go to images, and now I can actually simply edit this with Nano Banana 2. For example, I can say make the car a convertible. And you'll actually be surprised just how quick Nano Banana 2

is. Here you can see it generated that images in just a few seconds. Of course, if it is something that is a little bit different and you're not using a standard image, you can essentially prompt things into existence. Let's actually decide to prompt something from scratch. And considering we want more than one variation, let's go ahead and bump that up to three. Let's keep the aspect ratio the same and let's hit enter. It's important to know that images actually do generate in parallel, so there's no queuing, meaning that you can generate multiple images fairly quickly. So now you can see here, we do have multiple different images, and this is really useful for ideation depending on exactly what we want to do. Now when you click the images, you're able to see the quality, exactly what you want. And now, another thing that I really do like about this is that you can actually download these images, but it gives you three different options. You can opt for the original size, or you can choose to upscale these to get super high quality. Let's choose for 4K, and bear in mind when you're upscaling, don't schedule multiple upscaling jobs because for some reason that tends to mess up the system. You'll hear more on that later. But here you can see, we have the image in 4K. This is a lot better in terms of the detail, and this is something that we

could definitely use depending on the project that we are making. Now, of course, this image that I generated, there may be a few things wrong with it that I didn't initially like. So once you do generate an image, every time you click an image, it will pop up with this box asking you what you want to change. Rather than actually just prompting it immediately, which you can do. For example, I can say change the color of the car to purple. And so here you can see we managed to make that change. Now, of course, that is pretty easy to do, but sometimes there are specific things in the image that we want to place, but we're going to need a way to do that. So let's say for example, we want this taxi cab here to be a separate vehicle or we want to remove it entirely. What we can do is we can highlight this, and we can say remove the vehicle. And now it's going to use that location mapping to remove said vehicle. Here you can see that it actually removed the object in the exact location specified. Now this can be super useful when you want to remove something in a specific location, but you don't know exactly how to describe it. Or even if you want to add things. So let's say for example, now that we've removed this, of course, we

could have just replaced it. If we want to add something, let's say we want to add a small detail like a pothole here, we can actually do that. So I'm going to say add a pothole, and here you can see it actually adds a pothole in the exact location specified. And this is of course the kind of thing that we want when we're using the location in order to be able to edit something. Don't forget when you're editing images, if you really want to get precise, you can click the box icon twice, then the lasso tool will pop up, and then for example, you can really use your mouse and pretty much edit absolutely anything. Just a quick tip that you may have missed. Now that this essentially covers the bases of your images, additionally, don't forget that if you want even more control, what you can do is use the draw feature. So for example, on the left-hand side, you can see that there is a drawing feature. You can change this in terms of your pen size, and you can draw different things. Now, I'm not too artistic myself, but let's say for example, I want to add something that's a specific shape, what I can do is I can actually draw that out. So let's say for example, I want a table in the middle of the road. I know this is a very weird request, but let's just do it. I can actually draw a table here. So let me

actually draw that table. And I'm going to say add a table like the sketch. Now of course, you can see that adding with a sketch is less accurate. I did test this feature multiple times, so it probably isn't worthwhile you're using, but if you're more artistic than me, and you can figure it out, maybe it will be worthwhile your time. Now of course, a thing to remember is that don't forget you need to actually add these to your projects. So click save to project, and then it will be actually added. Otherwise, when you back out, you won't directly see the image right there. You'll actually have to click that same image, go back into it, look up at the history, and then download whatever specific image you do want. This is something that is a little bit tedious, so don't forget to save the images to project once you're done. Now, let's start to create a video. So let's say for example, we want to actually create a video. Pick the image, and then of course, let's animate this. So let's enter my prompt, which is car drives off cinematic view, and then I'm just going to input that in. Now, I've purposefully chosen to use VO3.1 fast because this one is also going to generate the fastest and it's going to use the least credits. Now, personally, when I'm actually generating multiple videos and ideating on a project, I will bump that

up to three times because it allows me to ideate faster and it not only does that, it also allows me to basically pick and choose which ones I want because I'm avoiding getting through hallucinations. Of course, if you don't have much credits, it is best just to try and iron out your prompt as best as possible, but do remember there is a noticeable difference between VO3.1 and VO3.1 quality cuz of course VO3.1 quality is more compute intensive, meaning that it renders it with higher effort, which means that the output is going to be of a higher quality, there will be less hallucinations, but also that will cost more credits. So, if you're doing something that does need better physics or better motion, I would pick better quality, but if you're doing something that is very simple and very straightforward that the AI shouldn't struggle with, VO3.1 fast is going to be your best bet. Let's say for example, you generate a video and you want to quickly see what it looks like. You can of course first click the video and view it in its full screen. Now, here you can see that this video didn't really generate in the way that I wanted. It was okay-ish, but of course you can clearly see that the video just did some

strange U-turn. Now, this of course isn't great. So, what you want to do is you want to just click this button and you can essentially regenerate this. Here, what I'll actually do is I'll actually switch this to VO3.1 quality. Remember, this will take a bit more time, but it should result in a higher fidelity clip. Now, of course, considering that I do have multiple credits, I'll actually generate three in VO3.1 fast just to show you guys how each variation can differ even if they're the same prompt. So, this is what the clip looks like with a VO3.1 quality. This, I think, is a much better rendition of what the car should actually look like, but let me show you guys exactly what happens when you use VO3.1 fast. You can see here that if we look at this first clip, the car is actually sliding around, which is rather unfortunate. Then if we look at the second clip, this is actually a much better rendition. So, you can see here that the car actually pulls off and takes a first, you know, right turn, which is pretty good. And then on the third clip here, let's actually see what it managed to generate for us. You can see that this car actually once again it did a 360 and then pulled off any other direction. Now, of course, like I said, your prompt will determine how the AI inputs and outputs certain videos. So,

it's important to note that the more detailed your prompt, the better of an output you're going to get. But another thing that is important to note is that if you want to review videos and you don't want to keep jumping in and out, you can just hit the hover button and you'll be able to see every single video play in a loop at your request. And this is super super useful when you're generating and ideating multiple clips. For example, for me, when I generate three clips, I don't want to jump in and out of each one. I want to see which ones were generated looking the right way and which ones absolutely messed up. Now, of course, let's say you want even more control than this. What you can do is you can use the start and end frames in order to create something that is a little bit more better. So, let's actually take a first image of the car. Then let's actually enter a prompt and let's say the same car but a closer view of the wheel. And let's click enter. And do remember that you can generate multiple images. So, here I'm actually going to say a closer view of the yellow car's number plates. And of course, if I want to, I can, you know, generate another angle. I can say a closer view of the wing mirror. You can see here that this image actually generated a hallucination. It actually thinks the number plate is yellow when it is actually white. So, what I'm going to do here is I'm actually going to change the

number plate to white. So, I'm going to say change the number plate to white. And here we can see the updated version, which I will now add to my project. So, now what we'll do is we'll essentially try and generate a video with more control. So, if we take this first clip right here and let's click animate. And then if we then take this and we click add to prompt, you can then see that we have the first frame and then the second frame. Now, it's important to note that if you want to view these images at a glance, just hover over them. Don't click them. I've made that mistake before. It simply just deletes them and you'd have to import them again. So, just hover over it and make sure that the clip is right. So, the first clip we have here is a McLaren. Second clip we have is a number plate. And what I'm going to do is I'm going to say slow cinematic and I'm only going to generate one for this. And considering that I have the first frame and the second frame, it shouldn't have an issue. And so here, you can see that the final video actually shows us what we want to see. You can see that it then pans into the number plate and it zooms right on it. So, that is pretty good. I probably should have changed the GB sign because it does morph in there, but of course, that is something that completely change. Now, of course, that is how you want to use the first frame {slash} end frame because it is something that is

super super useful. Now, let's say of course you want to change things even more. You can see you've got the camera action right here and this is where you've got multiple different things. When you click this, you're able to essentially set the camera motion and of course the camera position. So, this is something that is super super useful and a little bit more advanced if you know exactly what you're doing. So, for example, the dolly in {slash} dolly out, all of these different things show you exactly what it will look like based on the current horse. So, I think one of the easiest ones to show you is of course the orbit up. So, if we click orbit up here and we just click enter, it's going to take that last frame and then it's going to orbit up out of that final shot, extending the final video. Here, if you click the I icon, you'll be able to see just how many credits this actually cost. And this is a super useful feature from Flow because it allows you to edit your videos in a way that is much easier than thinking about the camera angle in your head or thinking about entire directional scenes, which is pretty difficult to manage. And so here you can see it actually gives us the up pan there. And of course, that takes the first frame. So, do bear that in mind when you're generating these clips. But like I said,

if you actually do see on the left-hand side, these characters are actually morphing. So, do bear that in mind when generating with a VO3.1 fast, the physics of course isn't that good. In the future, this will likely change to another model, but just bear in mind, the lower the model is, the less quality it's going to have. Now, of course, with videos, if you actually want to get higher quality, what you can actually do is upscale them. Standard upscaling is essentially free, but 4K upscaling is around 50 credits and will take around 2 minutes to do so. So, do bear that in mind. Now, for example, let's say you actually wanted to edit something in this video, you pretty much can. So, let's say we click insert. What we can do is we can actually insert a butterfly flying. So, let's say add a butterfly flying across the screen. And so here you can see it actually did generate the clip, but of course, this isn't exactly the highest quality. Do remember that I am using VO3.1 fast, but it will show you exactly what to do once the better models do come out. And of course, if you do want to change that model, just don't forget to click here. And then when you're doing your insert and stuff, it will essentially do that. Additionally, it will work the same. So, if you do want to remove that butterfly, you can also do that. Now, let's say for

example, this car was annoying me. I can literally say remove this car. And of course, this car was present in the last video. So, now let's remove it see what happens. And here you can see in the final clip, it actually removed the car mostly from the generation. It does actually appear in the final third, but for the most part, it does actually disappear. Now, of course, if you do ever want to extend your clips, you can literally just click extend. And depending on what you want to happen, you can of course direct the scene in whichever way you do want. really that creative, so I don't know what to put here, but it is still a very useful feature. Now, something that is also useful, as you're scrolling through the timeline, maybe you'll see something that you like as a specific frame. So, for example, let's say we want this frame from our video. What you can do, in the top right, if you click this image, you'll see the save frame icon. Here, essentially all you'll need to do is save that frame. And then when you go back to your images, you'll see that image right there. This is super useful if you're trying to ideate, you're trying to then animate from here, you're trying to save this. It is just super useful to have on hand rather than having to go in and screenshot. Another thing that is pretty useful when you're working on your projects is navigation.

Now, as time goes on, your projects will likely get bigger and bigger, more and more complex, and you'll have trouble looking through the different things. So, what I would say is make sure that you're naming things as you're going through. For example, right here, I'm actually going to rename this as McLaren butterfly one. And then what I'll do is I'll click save. Then what I'll do here is I'll name this one McLaren butterfly two, butterfly two. And the reason you want to do that is because when your projects get to really really big, as it has with me many times before, you'll struggle to find a specific clip. So, if you want to find that specific clip, go to the top, then type in McLaren or just type in butterfly, and then you'll essentially see your clip there. Now, what's really cool about this is that it auto generates names for you. So, essentially if you haven't named everything, it should name things for you. So, for example, you can see the convertible right down there. If I type in convertible, you'll see that it actually does come up. So, don't worry about it too much, but it is something that if you're trying to navigate the project, it will be tremendously useful and save you tons of time. Now, another quick feature that Google have added just to make navigation that much better is that if you actually put the at sign

in, this window will pop up and then you can essentially add different things in. For example, I can add in this McLaren specifically and then prompt away on whatever I want. Now, you might be wondering why is this specifically useful? Well, the reason that this is specifically useful is I'll show you that it's basically designed for interacting with certain characters and certain objects. Let me show you guys an example. So, I'm going to start a new project here and here I'm just going to make something completely fantasy up, but I's going to show you exactly why the app is so great. So, what I'm going to do is I'm going to say uh yellow orc. And I'm going to say white background character design. And then what I'm going to do is I'm going to render an orc's car character design 3D render. And so here you can see I've actually generated a few assets. All I did was just put, you know, a yellow orc character render and I made sure to include that these have white backgrounds. Now, the reason you do this is because let's say you wanted to actually build up a landscape or build up a kind of, you know, character story. What you want to do is you want to rename all of this. So, let's say this character's name was like Michael or something. I don't know what orcs are called, but let's say we name it Michael. Then let's say this is the chain ball weapon. I'm just going to

name this um let's say the, you know, the the chain weapon, okay? And let's say that this is of course the vehicle. Uh we could just say this is the mega truck, okay? So, now if you want to prompt this rather than, you know, dragging everything in, you know, step by step, what you can just do is you can say, you know, at Michael and then you could just say Michael driving is, you know, and then at and then put mega truck and then just grab the mega truck and then in and then of course, if we want to do the fantasy landscape, we can do at and then fantasy medieval landscape. So, now you can see I have Michael driving his mega truck in this fantasy landscape. And now it's going to actually use that image as the driving image. And then, so now we can do is we can then put this enter. And then, so yeah, you can see that the final picture that we get is here. And it's probably going to still work out the same way if you prompt it in another way. But with this, I just think it's much easier to prompt using the app because when you have a big project and you're trying to, you know, iterate multiple multiple times, this new feature that they've added just makes things 10 times easier. Additionally, if you want to filter it even more, you can filter it by the oldest or newest. You can filter it by your favorites. That's if you click the

like button. Or you can filter it by images, videos, or different aspect ratios. There's multiple different options. So, this is something that is once again super useful. In addition, if you want to view the project in a different way, you click the settings icon. And then, this is where it allows you to view them in different ways. So, for example, let's say these grids are a bit too big. I can literally click S and this is going to put things into a much smaller way that we actually see all of my projects at scale. If you only have a few things, I can just click L. It's going to put them into much larger. Medium size is best for me. If you Now, if you want to actually view things as you've done in the batches, you kind of click batch here. And then, it's going to show you every single clip that you generated within that specific batch. This is going to be useful for various different things. But for me, I just like to leave them on the grid because it's super useful. Of course, something that could be useful to you is if you want to show the tile details on. This basically is just where they're going to show you the text right here. So, you can click it off. There's no text. If you click it on, if you want to have the text there, just literally hover over it. And of course, sound hover on or off is basically where you're hovering over videos, it will either turn the sound on or off. If you want even more navigation, if you go on the left-hand

side, the first one is basically everything at a glance. The second one is going to be your images. The third one is going to be your videos, all of them. The fourth one is going to be anyone that you've clicked the like button on. So, make sure that you have some favorites there. And then, of course, the last one is going to be your uploads, things that you actually put into the project from your device. You can see I click plus. I can actually add multiple different things into that video. So, you can see I can add the McLaren. I can add the wheel. I can add the car driving here. And I can say McLaren cinematic video. Then I can go ahead and click generate. And if you want to search between your assets, don't forget that you can literally search between all of your other projects here. For example, I can see my other project here. And I can see my other stuff here. So, here you can see the final video. And here on the right-hand side, you're able to see all of these small things that I was using to generate this cinematic clip. So, I think this is pretty good. And of course, ingredients is essentially where you're using multiple different images as a reference for your final video. I think it's a lot more useful when you're trying to create a sort of background slash scene rather than working with specific subjects because it can get a little bit messy. Now, of course, once you've generated videos and images,

let's actually go to the scene builder if we want to build something cool. So, on any video clip, what you can do is you can literally click the top right. Then you can click add to scene. And then, essentially, what it will do is it will pop up with a scene builder. And then, you can literally click switch to scene builder or you can just click the top icon in the top right. And so, now what this does is this is probably where you can start building your filmmaking scenes. So, for example, this is a really nice clip. I like this one. This is a super, you know, sort of cinematic clip. So, then what I'll do is I'll add more clips to this. Let's add this scene to the clip. Let's add this to the scene builder. And if we click the scene builder icon right here, you can see that essentially, what it does is it automatically pulls all the clips together. Now, do remember that the scene builder is essentially the final thing where you tie all the clips together. This is not the place where you start screenshotting and editing everything together. That is, of course, what you do before. So, if you wanted to create a cohesive scene, the scene builder isn't really the place to do that. It's just where you merge all the clips together, basically like a timeline.