Transcription
Google's V3 model is now in N810, which means that the AI videos you may be seeing blow up on socials, like this Yeti vlogger getting tens of millions of views, or this Filipino grandma that's also getting popular on Tik Tok, which are all generated by Google's V3 model, which shook the world when it was announced just a few days ago because of how realistic the video generations are looking, can now be automated through N8N and without Google's $200 per month price tag.
And in case you're not aware, N8N is a no-code automation tool quite similar to Zapier or make.com, if you know those platforms. But you can see it shot up in demand quite recently because of how flexible it is with implementing a lot of these new AI technologies. And the good news is anyone can do it, even if you're a complete beginner, mainly due to its visual interface, which we'll be going through in this video.
So, if you put these two together, N8N and V3, it will definitely be a very valuable skill set that you can offer for your content creator business clients or use on your own if you're an aspiring content creator yourself. So, in this video, I'll be sharing with you this N8N AI agent template, which, if we run it, basically what it does is it will craft prompts tailored for V3. It will then automatically pass that to V3's AI model to create videos like these on autopilot.
For all my followers, remember, never eat yellow snow unless you like surprises. We cut to commercial, duh. And that's, I mean, that's as realistic as you can get. And if you have this system set up once and you have an account like this, you can see that they already have 200,000 plus followers, even if they only have something like 11 videos on here. So, if you set up this workflow once and let it run in the background, then you can see how powerful this automation can become. And if you watch till the end, I'll also share with you how you can autopublish these videos to any social media channel of your choice.
And it's probably a good idea to invest some time learning about these new tools, especially if you just check on Google Trends just how much demand is forming within these new platforms. So, if you have some time today, then watch this video and let yourself learn a new skill that can put you at the very edge of what's possible now with AI.
By the way, if you're new here, my name is Jay, and our mission here at Robo Nuggets is to make AI easy to learn for anyone who wants to, no matter what your technical background is. We run RoboLabs, which is our AI agency arm, helping businesses get to the future with AI. And we also operate one of the largest communities of AI practitioners globally. We now have something like 1500 plus members across the globe.
And in fact, to make this as easy and accessible as possible to anyone who wants to learn it, we'll be uploading this N810 template, which you can just import into this tool to instantly set up this workflow, which you can get access to just in the link in the video description. And it also comes with this quick instruction manual on how to set it up. But just so you know how it works, let's dive into it because really, if you break down this workflow, there's essentially just three layers to it.
This whole piece at the top are basically nodes where we've set up our own AI agents that are creating the prompts that's tailored for V03. Then we have this next layer where we pass along that prompt into V03's image model, which we'll go through in a bit. And then we have this final Google Sheets node that just logs everything so that we don't lose all the videos that we are generating.
So if we open this up one by one, basically the trigger that we set here is a schedule trigger, and you can obviously change this if you want, if you want it to run daily or every hour, that's fine too. But the first AI agent that you have here is an ideas AI agent, which, if you just open that up, you'll see that for this AI agent, we are giving it this instruction or prompt where the task is for it to give us an idea about this specific niche that we are interested in. So for this example, it's basically the Yeti vlogger that we saw from earlier.
But more importantly, if this is your user prompt, kind of like what you send to chat, it's also important to define what your system prompt is. And a system prompt is basically an AI agent's role in life. So if we open that up, you'll see that this is just a pretty long system prompt that is guiding this AI agent on what's the output that it should provide depending on what the user is asking for. So you can read on this on your own time, but it just defines the style, the output format, and different things that you can tweak on your own.
But when a workflow runs and calls on this AI agent, basically it will look at this user message and generate this output, which is already organized depending on the system message that we just crafted. So you will have a caption that's fit for social channels. You would have the idea, which is important because that is what we'll pass along to the next agent, which you will see in a bit, and an environment just to define what setting our video will be on. And then this status for production is just important for logging purposes because if we click away, you'll see that the next node we have is just a Google sheet node, which, if you open that up, you'll find that it's connected to this Google sheet template where basically N just loaded the idea, the caption, that status, which we set as for production, as well as this environment prompt, just so that we have proper tracking of each idea that our ideas agent generates.
Now, if we go back to our workflow, you can see that the next step is this prompts AI agent. And if we open that up, you'll see that we have a simple user prompt here, which, if we maximize that, it's essentially asking the agent to provide a V3 prompt for the idea that our previous AI agent generated, which for this one would be this, where the Yeti is vlogging. And we also pass along that environment to give it some guidance on what setting our character is in. But even more important than that is this AI agent system prompt so that it knows what to do when you call on it.
So if I expand this, this is just a system prompt specific for V3, and you can edit this as you wish, but it just provides some guidance on what are some essentials for the final prompt that it will generate for us so that we can automatically pass along a fairly good prompt to V3 every time we use the model. So you can see the output here is this pretty well-detailed prompt that follows our instructions here while also considering the fresh idea that our previous AI agent generated.
And there, this is actually your end in mind for this first row of nodes. Because now that you have the final video prompt generated, you can then pass that along to V3 through this node, which is an HTTP request. And if in case you haven't used this node before, it's essentially N8N's way to call on third-party tools, including this one that we'll be requesting from to use V03. And you'll see we just have the method here as post because we are posting a request, similar to how you would post a mail. And the service that we are requesting from is this service called file.ai. AI, specifically their V03 model instance.
And if you haven't used file.ai before, it's basically a way for you to call on several AI models using just one platform. So an analogy I gave before is if you're trying to order from a restaurant, you can of course directly contact that restaurant. But if you have something like Uber Eats, which centralizes everything, it's sometimes much easier to just go to Uber Eats and request the service of those restaurants from that one platform. So in that analogy, file.AI is Uber Eats. And instead of you figuring out how to connect to V03 and so on, they have a lot of models here already centralized. So in that way, if Uber Eats is for restaurants, then file.AI is for AI models. And you can see that literally just hours ago, they announced the V3 integration. And they were the first to do so as well. And I myself first got to know that it's available because a member of our community posted in our group that they sent this email just a few hours ago announcing that V3 is now available in their platform.
So anyway, back to our workflow. We're basically using file.AI in this case. And we also have our credentials set up through a pass key that you can get once you sign up to file.AI. And then you can see the request that we are passing on to them is very simple. It's just this one line where we're passing along the prompt, which, if we expand that, what it's doing is it's just dynamically getting the prompt from our AI agent and posts that as a request to file.AI. And once this step has completed, it will say that the request is in Q, which basically means that V3 is now hard at work with generating that video.
And so the next step is to literally wait for V3 to finish generating that video. And so if we expand this wait node, you'll see that we just have a set amount of time here, which now we set as around 10 minutes. So that it gives ample time for V3 to generate the video. And then once that time is over, we just have this second to the last node, which is another HTTP request. But this time it's just a get request because all we're doing is asking file.AI for the status of that video generation job that we just posted to it, which is being passed along through this request ID that we obtained coming from the post request that we had previously. And so when this node was executed, you'll see that we now have this video URL, which, if you paste to any browser, you'll be able to preview what the final output is.
For all my followers, remember, never eat yellow snow unless you like surprises. We cut the commercial, duh. So there, that's looking pretty realistic. And so the final step here is just another Google Sheets note, which, if we open that, that just connects to our Google Sheet template once again, where all we are doing is finding that row that we need to update and then marking the production status as done, and then putting that URL of the video under this final output column so that we make sure that we have it properly logged. And when that node runs, it'll pretty much look like this where you have the production status now set as done. You have the and you have the final output of the video added to your tracking sheet as well.
So now if in case you want to auto post this video into any social platform that you want, I can recommend to you this tutorial where we went through a tool that can enable that, which, if you're part of the RoboNuggets community, we also have a discount code for that that gives you 30% off for life. So make sure to use that if in case you plan to go that route.
But there, very quick intro to how to use V3 and N8N, which we now can do thanks to File AI's integration. Now this workflow template I'll be sharing and we have here some step-by-step instructions also on how to set it up for your own space. But what's very important to mention is the cost of V3 Tree because if we go back to File A's V3 page, you'll see that the cost here is not cheap at all. So, the summary cost for each video, which is always 8 seconds for V3 at the time of this recording, it's going to be $6 per video. And so, that's obviously very high and probably not meant for everyone. So, that's just something to keep in mind if you're testing this out.
That said, $6 is still cheaper than Google's paid plans. So, in some ways, it does make V3 more accessible to people. But, that's obviously the cost now with a lot of these technologies like cloud storage, like even electricity. A year from now, we'll likely look back in this video and see how the costs of these models are so high in this point in history because obviously the technology is literally just a few days old. But if you wait a bit, most likely the costs will go down as well. But at least now you're equipped with the skills you need to take advantage of these new developments in the AI space.
So there, I hope you learned something new from this quick video. And if in case you're not yet part of the Roboon Nuggets community, you can check that out in the link in the video description if you prefer to join us. The community is basically this AI education program that we run where we have a lot more lessons on artificial intelligence along with all the templates that you would need to get you started. We also have a lot of events that are happening every week. In fact, we're about to do global meetups soon as well, if that's something you're keen to do, or if you're looking for partners or paid opportunities. Then a lot of our members, including myself, also post opportunities and partnership collabs here if that's something that you are looking for.
So there, no pressure. Check that out if that's for you. And I'll see you all next time. Thank you. [Music]