<b>Google's Gemini 3.0 is</b>

<b>finally here and it's incredible.</b>

<b>And in this video I'll show you how you</b>

<b>can start using Gemini 3.0 in N8N today.</b>

<b>In the next few minutes we'll go over</b>

<b>everything from getting your API key</b>

<b>to using the impressive</b>

<b>features offered by Gemini 3.0 Pro,</b>

<b>which includes not only text generation</b>

<b>but also analyzing</b>

<b>images, audio and video.</b>

<b>Enough talk, let's jump in.</b>

<b>I'm not going to spend too much time</b>

<b>going over the performance of the model,</b>

<b>as I'm sure all of you have already seen</b>

<b>these performance charts.</b>

<b>Gemini 3.0 Pro is outperforming all these</b>

<b>other flagship models in a lot of</b>

<b>different instances.</b>

<b>The model also has a massive context</b>

<b>window at 1 million tokens.</b>

<b>And in terms of pricing, it's about $2 if</b>

<b>you stay within the 200,000 token limit</b>

<b>and about $4 if you</b>

<b>exceed the 200k token window.</b>

<b>Now let's have a look at</b>

<b>using Gemini 3.0 Pro in N8N.</b>

<b>The first thing we need</b>

<b>to do is get our API key.</b>

<b>To do that, go to aistudio.google.com.</b>

<b>So by the way, what's really cool is you</b>

<b>can try Gemini in this interface as well.</b>

<b>And you can try Gemini 3 for yourself.</b>

<b>Now to get our API key, what we have to</b>

<b>do is click on "Get API Key".</b>

<b>Then on this page,</b>

<b>click on "Create API Key".</b>

<b>Then let's give it a name.</b>

<b>I'll just call it "N8N Tutorial".</b>

<b>Then from the list of projects, we have</b>

<b>to select a Google</b>

<b>Cloud Platform project.</b>

<b>Now if you don't yet have one, simply</b>

<b>click on "Create Project".</b>

<b>Then just give your project a name.</b>

<b>And then select your project from the</b>

<b>list and click on "Create Key".</b>

<b>Right, so I've just created this key.</b>

<b>And you won't notice</b>

<b>it on the quote tier.</b>

<b>I'm currently on tier one.</b>

<b>For you, it might show "Set up billing".</b>

<b>Now just like with all these leading</b>

<b>models, this is a paid service.</b>

<b>So what you have to do is click on "Set</b>

<b>up billing" and then</b>

<b>follow the instructions.</b>

<b>Now it's worth mentioning that when I set</b>

<b>up my billing account,</b>

<b>Google actually gave me a lot of free</b>

<b>credits to play with.</b>

<b>So chances are you</b>

<b>won't pay anything for this.</b>

<b>All right, and we'll revisit this page</b>

<b>when we set up our credentials in N8N.</b>

<b>Now I do recommend updating your N8N</b>

<b>instance to the latest version.</b>

<b>If you're using N8N Cloud, then you</b>

<b>should already be on the latest version.</b>

<b>But by the way, while we're here, N8N</b>

<b>Cloud is really not cheap.</b>

<b>And personally, I like to self-host my</b>

<b>N8N instances, which is way cheaper.</b>

<b>In fact, hosting your currency has a</b>

<b>Black Friday special,</b>

<b>where you can self-host your N8N instance</b>

<b>for as little as $5 per month.</b>

<b>Comparing that to the 20 euros per month</b>

<b>on N8N Cloud, this is a steal.</b>

<b>So all you have to do</b>

<b>is select your instance.</b>

<b>I personally like KMv2,</b>

<b>so I choose your plan.</b>

<b>Then you can select your term.</b>

<b>Obviously, with 24 months,</b>

<b>you'll get the best deal.</b>

<b>But if you just want to go month to</b>

<b>month, this is still only $10.</b>

<b>Then scrolling down,</b>

<b>select N8N under Applications.</b>

<b>Click on Coupon Code</b>

<b>and enter the code Leon.</b>

<b>This will apply an</b>

<b>additional 10% discount.</b>

<b>So let's continue to checkout.</b>

<b>You will then have</b>

<b>access to this dashboard.</b>

<b>All we have to do is click on Manage App.</b>

<b>So this can wait and complete this form.</b>

<b>Then I'm just going to skip this pop-up.</b>

<b>And if you receive this screen,</b>

<b>definitely send yourself a license key.</b>

<b>Then get your email and copy that key.</b>

<b>And then back in N8N, go to Settings,</b>

<b>click on Interactivation Key.</b>

<b>This will unlock a bunch of</b>

<b>very cool features in N8N.</b>

<b>So if you want to save some money,</b>

<b>definitely use the</b>

<b>link in the description</b>

<b>to set up your N8N instance on Hostinger.</b>

<b>Right, so from the N8N dashboard, let's</b>

<b>create a new workflow.</b>

<b>Let's rename this to Gemini 3.0 Tutorial.</b>

<b>Cool.</b>

<b>I'll simply add a trigger manually node.</b>

<b>And then let's add a new node and let's</b>

<b>search for Google Gemini.</b>

<b>Now we have access to a</b>

<b>whole bunch of different actions</b>

<b>like analyzing audio,</b>

<b>analyzing documents,</b>

<b>analyzing images, analyzing</b>

<b>video, and much, much more.</b>

<b>Let's start with the fundamentals.</b>

<b>Let's click on Message a Model.</b>

<b>Now the first thing we need to do is set</b>

<b>up our Gemini 3.0 credentials.</b>

<b>So on the credentials,</b>

<b>let's click on the dropdown.</b>

<b>Then let's click on</b>

<b>Create New Credential.</b>

<b>I'll just give this a</b>

<b>name like Gemini Tutorial.</b>

<b>Now this is where we</b>

<b>have to add that API key.</b>

<b>So of course, let's go</b>

<b>to our Google AI Studio.</b>

<b>Let's copy the key that we just created</b>

<b>and paste it into this field.</b>

<b>Let's click on Save.</b>

<b>And our connection was successful.</b>

<b>So under the Model list, you can see a</b>

<b>large list of available models.</b>

<b>Now of course, we are</b>

<b>interested in Gemini 3.</b>

<b>So I'm actually just</b>

<b>going to search for Gemini-3.</b>

<b>And now we can see Gemini 3 Pro.</b>

<b>And now we can interact</b>

<b>with it like any other model.</b>

<b>So let's just say,</b>

<b>hey, why is the sky blue?</b>

<b>And if everything was set up correctly,</b>

<b>we will receive a response, which we do.</b>

<b>Now the Gemini 3.0 Pro model</b>

<b>already includes reasoning.</b>

<b>And at the moment, there's really no way</b>

<b>to change the reasoning effort in N8N.</b>

<b>But of course, knowing N8N, they'll</b>

<b>probably release an update</b>

<b>by the time you watch this video.</b>

<b>And hopefully by the</b>

<b>time you watch this video,</b>

<b>there will be some kind</b>

<b>of reasoning effort option.</b>

<b>Now besides for generating text, Gemini</b>

<b>3.0 is a multimodal model.</b>

<b>So it's able to analyze</b>

<b>audio, video, and images.</b>

<b>Let's see that in action.</b>

<b>So let's have a look at Gemini 3.0's</b>

<b>ability to analyze images.</b>

<b>So again, let's add our Gemini node.</b>

<b>And under actions, let's</b>

<b>select analyze an image.</b>

<b>Let's select our credentials.</b>

<b>And from the model list,</b>

<b>let's select Gemini 3 Pro.</b>

<b>And for the text, we'll just leave it on</b>

<b>what's in this image.</b>

<b>And under input type, we can either</b>

<b>provide a URL to the image</b>

<b>or we can upload it as a binary file.</b>

<b>Let's have a look at both options.</b>

<b>So for the image, I'll</b>

<b>just go to Pexels.com.</b>

<b>And let's actually</b>

<b>grab this image over here.</b>

<b>So I'll just right click on this image.</b>

<b>Let's copy the image link.</b>

<b>And let's paste that</b>

<b>into the URLs field in N8N.</b>

<b>Let's run this.</b>

<b>Based on the visual elements, this image</b>

<b>depicts a joyful scene</b>

<b>from a traditional Vietnamese wedding or</b>

<b>engagement ceremony.</b>

<b>If anyone watching this is from Vietnam,</b>

<b>I would love to know how</b>

<b>accurate this actually is.</b>

<b>Leave a note down in the comments.</b>

<b>Now of course, we don't have to use URLs.</b>

<b>We can also use any</b>

<b>binary files that were either</b>

<b>passed into the workflow or generated</b>

<b>within the workflow.</b>

<b>So to use binary files, what we can do</b>

<b>is get the file into</b>

<b>the workflow somehow.</b>

<b>So I'll just add a form.</b>

<b>And let's give it a name.</b>

<b>And for the elements,</b>

<b>let's just call this one image.</b>

<b>And for the element</b>

<b>type, I'll select a file.</b>

<b>Just need a single file.</b>

<b>And yes, this is a required field.</b>

<b>So let's execute the step.</b>

<b>Then let's download an</b>

<b>image like this one over here.</b>

<b>I'll just go with the small size.</b>

<b>Then let's select that</b>

<b>image and submit the form.</b>

<b>Cool.</b>

<b>Now we have that file in</b>

<b>our workflow as binary data.</b>

<b>So all we really have to do</b>

<b>is connect these two nodes.</b>

<b>Then in the Gemini node,</b>

<b>let's change this from a URL</b>

<b>to a binary file.</b>

<b>And we just have to make</b>

<b>sure that this name matches</b>

<b>this name on the left,</b>

<b>which is actually called image.</b>

<b>All right, let's run this.</b>

<b>And based on the image, here is a</b>

<b>detailed description, subject</b>

<b>and action, a pregnant woman.</b>

<b>The central figure is a woman with dark</b>

<b>shoulder length here</b>

<b>standing in profile facing</b>

<b>the right side of the frame.</b>

<b>This is awesome.</b>

<b>Now let's also have a look at</b>

<b>its ability to analyze video.</b>

<b>So let's remove this node.</b>

<b>Then let's search for Gemini.</b>

<b>And in this list,</b>

<b>let's select analyze video.</b>

<b>Of course, let's select our credentials.</b>

<b>And from the model list,</b>

<b>let's search for Gemini 3 Pro.</b>

<b>As with analyzing images, we</b>

<b>can provide a URL to the video.</b>

<b>Or what we can simply</b>

<b>do is use a binary file.</b>

<b>Both options will work.</b>

<b>To demonstrate this,</b>

<b>let's just use pixels again.</b>

<b>So I'll search for a</b>

<b>video like this one over here.</b>

<b>And let's copy the video link.</b>

<b>It's basically into the URL</b>

<b>field and press execute step.</b>

<b>This video captures a serene moment</b>

<b>featuring a young woman</b>

<b>dressed in traditional</b>

<b>blue attire, likely Hanfu.</b>

<b>Standing in an ornate courtyard that</b>

<b>appears to be part of a temple</b>

<b>or historical site.</b>

<b>Again, let me know in the comments how</b>

<b>accurate this actually is.</b>

<b>But the description of the clothes and</b>

<b>the setting does seem correct to me.</b>

<b>So let's try another one actually.</b>

<b>So here's this video of this hippo.</b>

<b>I'm actually going to</b>

<b>download this video.</b>

<b>Then let's change our form submission.</b>

<b>So instead of it being an</b>

<b>image, let's upload a video.</b>

<b>Then I'll execute the form.</b>

<b>Let's select the video that we just</b>

<b>downloaded and submit this.</b>

<b>Cool.</b>

<b>Now that we've got this</b>

<b>video in the workflow,</b>

<b>let's open up Gemini again and let's</b>

<b>change the type from URL to binary file.</b>

<b>And we just have to make</b>

<b>sure that the binary name</b>

<b>is the same as the name on the left hand</b>

<b>side, which is video.</b>

<b>Let's run this.</b>

<b>The video shows a large hippopotamus</b>

<b>standing in profile on a concrete</b>

<b>or sandy bank within a zoo enclosure.</b>

<b>That is completely accurate.</b>

<b>Now Gemini 3 is also</b>

<b>really good at analyzing audio.</b>

<b>So let's add Gemini back in and under</b>

<b>actions, let's select analyze audio.</b>

<b>Let's select our credentials.</b>

<b>And for the model, of course, we'll</b>

<b>select Gemini 3 Pro.</b>

<b>Now we can provide a URL or of course we</b>

<b>can upload an audio file.</b>

<b>Now for fun, I downloaded the audio only</b>

<b>of this very famous clip.</b>

<b>So I'm not going to play the audio itself</b>

<b>out of fear of</b>

<b>getting a copyright strike,</b>

<b>but let's actually change the form.</b>

<b>So I'm going to change</b>

<b>the field name to audio.</b>

<b>Then let's run the form.</b>

<b>Let's upload that audio file and then</b>

<b>let's go back to analyze audio</b>

<b>and I'll just rename data to audio.</b>

<b>So basically in this audio</b>

<b>clip, there are two speakers.</b>

<b>So there's Sydney Sweeney saying</b>

<b>something about genes.</b>

<b>And by the end, there's a man's voice</b>

<b>that's doing a bit of narration as well.</b>

<b>So what we can do maybe is say, how many</b>

<b>speakers are in the audio clip?</b>

<b>And Gemini 3 is saying there are two</b>

<b>speakers in the audio</b>

<b>clip, which is correct.</b>

<b>Let's also add how many</b>

<b>male and how many female.</b>

<b>So it would be interesting to see if the</b>

<b>model can actually distinguish</b>

<b>between female and male voices.</b>

<b>Based on the audio clip, there are two</b>

<b>speakers, one female.</b>

<b>And that's interesting.</b>

<b>It actually knows that it's Sydney</b>

<b>Sweeney and a male voice.</b>

<b>That is really interesting.</b>

<b>And finally, let's ask what</b>

<b>is the audio really about?</b>

<b>The audio is a play on words based on the</b>

<b>identical pronunciation,</b>

<b>afterward genes, and genes.</b>

<b>How cool is that?</b>

<b>Finally, we can also use</b>

<b>this model to build AI agents.</b>

<b>So let's select the on chat message</b>

<b>trigger, and then let's add an AI agent.</b>

<b>Really, all we have to do</b>

<b>now is on the chat model,</b>

<b>we can search for the</b>

<b>Google Gemini chat model.</b>

<b>Then let's select our credentials.</b>

<b>And under the models,</b>

<b>let's search for Gemini 3 Pro.</b>

<b>And that's really it.</b>

<b>You can add memory, you can add tools.</b>

<b>This model is really</b>

<b>good at doing tool calling.</b>

<b>But let's just see if this works.</b>

<b>And it does.</b>

<b>We get our response back.</b>

<b>Now that you know how to add Gemini 3.0</b>

<b>Pro to your in-at-end workflows,</b>

<b>go and build something awesome.</b>

<b>I'll see you in the next one.</b>