WEBVTT 00:00.280 --> 00:05.920 In this video, we'll be talking about how we can connect perplexity with Nadine to make your research 00:05.920 --> 00:07.160 ten times better. 00:07.160 --> 00:10.520 So we'll be setting up perplexity with our A agent. 00:10.560 --> 00:15.960 I will show you how to set up perplexity node and even how to set up HTTP request. 00:16.160 --> 00:19.280 And what we'll set up in this video simply looks like this. 00:19.280 --> 00:23.520 So we've got firstly a agent that connects to perplexity. 00:23.520 --> 00:28.600 And here OpenAI model, there the perplexity node and HTTP request. 00:28.800 --> 00:31.880 So firstly maybe let's start with a agent. 00:33.080 --> 00:37.760 So to begin with let's just pick our trigger which will be on chat message. 00:38.000 --> 00:41.040 So whenever we've got the message we trigger this automation. 00:41.040 --> 00:43.760 And right away we connect it with a agent. 00:44.760 --> 00:46.280 So that's truly amazing. 00:46.280 --> 00:52.680 We can connect our agent with ChatGPT model and also perplexity. 00:52.960 --> 00:59.080 And this is the combination that allows us to make really professional and highly interesting content 00:59.120 --> 01:01.480 based on the best research available. 01:01.520 --> 01:07.370 Because we need to remember I'm actually perplexity is the best tool out there on the internet to make 01:07.410 --> 01:08.010 a research. 01:08.010 --> 01:15.970 And finally, we've got the opportunity to use perplexity, because actually today at June 13th, um, 01:16.010 --> 01:16.850 2025. 01:16.850 --> 01:21.810 So if you are watching this video right now, perplexity, insight and then was released just one day 01:21.810 --> 01:22.690 ago or so. 01:22.890 --> 01:29.130 So yeah, it's pretty new tool and I think it will gain a lot of popularity, um, within the weeks, 01:29.130 --> 01:31.690 months or even in long term. 01:31.730 --> 01:34.090 But now let's set up our chat model. 01:34.090 --> 01:42.570 So for this case OpenAI, we can choose GPT four or mini, which is really affordable and pretty effective. 01:42.570 --> 01:44.290 So for this case GPT four mini. 01:44.850 --> 01:50.410 And now when it comes to the tools, we can click on this plus and add perplexity. 01:51.330 --> 01:53.650 So we see this perplexity tool. 01:53.810 --> 01:57.570 And remember actually it's available with the newest update. 01:57.570 --> 02:02.330 So if you can see that probably you don't have the update like in your cloud. 02:02.570 --> 02:05.490 Nevertheless let's click on the perplexity tool. 02:06.180 --> 02:09.140 And then we need to connect our credential. 02:09.140 --> 02:11.260 So let's click on Create New Credential. 02:11.260 --> 02:15.260 And now we need to provide our API key to get your API key. 02:15.300 --> 02:17.300 Move on to perplexity I. 02:17.580 --> 02:22.460 And in the left bottom corner click on an account and click on my account. 02:22.860 --> 02:26.020 After you do this here you should see API keys. 02:26.060 --> 02:29.380 Click on it and then just click on Create key. 02:30.380 --> 02:31.900 After a while you've got your key. 02:31.940 --> 02:33.140 You can just copy it. 02:33.980 --> 02:38.100 Also remember here on clicking by clicking on these three dots, you can just delete it. 02:38.140 --> 02:40.260 And for sure after this video I will do this. 02:40.260 --> 02:44.500 So also remember about your security if you are for an example. 02:44.660 --> 02:51.300 Um, like yeah, I'm showing someone the automation or just you download the JSON file. 02:51.300 --> 02:54.500 So then you just get here your API key. 02:54.500 --> 02:55.860 So remember to remove it. 02:56.380 --> 02:57.500 Let's move on to Na. 02:57.500 --> 02:59.220 Then we can just paste it here. 02:59.500 --> 03:00.860 We can save it. 03:00.860 --> 03:03.060 And then we are ready to go. 03:03.100 --> 03:04.260 We can just close it. 03:04.940 --> 03:08.110 And now in general the last thing we need to do. 03:08.150 --> 03:09.870 We need to set up our text. 03:09.910 --> 03:14.070 We've got this option that allows us to ask AI to define this parameter. 03:14.070 --> 03:17.430 So we can click on it, and then we are ready to go. 03:17.870 --> 03:20.990 In general we can provide some message inside. 03:21.310 --> 03:21.550 Yeah. 03:21.590 --> 03:22.710 Inside our automation. 03:22.710 --> 03:29.990 So for an example what are five the best AI tools for video editing. 03:30.470 --> 03:31.710 And 2025. 03:32.030 --> 03:33.830 When I run it you see what happens. 03:33.870 --> 03:36.070 Actually it uses ChatGPT. 03:36.510 --> 03:42.990 Then it uses perplexity to do a research about what we asked, and then it will just go back to ChatGPT 03:42.990 --> 03:44.990 and it will create for us a response. 03:44.990 --> 03:52.390 So this combination is highly, I can say effective because from the one side we've got our ChatGPT, 03:52.430 --> 03:53.990 which is a huge LLM model. 03:53.990 --> 03:56.590 This is so actually, you know, you know, ChatGPT. 03:56.950 --> 04:01.350 And on the other side we've got perplexity that does the entire research. 04:01.350 --> 04:06.510 So you can combine these two tools and you will see what happens actually here. 04:06.630 --> 04:12.520 So we need to wait a few seconds more and hopefully the response will be created. 04:12.520 --> 04:17.080 So I will show you again how it performed. 04:17.920 --> 04:22.200 So after the process is done, I can click on a agent and here on logs. 04:22.440 --> 04:25.920 So we've got the first step which is OpenAI model. 04:26.160 --> 04:32.960 Then it came to perplexity and it asked, all right what are five the best tools for video editing and 04:32.960 --> 04:35.360 perplexity did the entire research. 04:35.560 --> 04:40.160 And after that we've got our response from ChatGPT. 04:40.200 --> 04:41.840 That's highly professional. 04:42.000 --> 04:45.240 And again, we connected perplexity with OpenAI. 04:45.920 --> 04:49.720 So this is how we connected perplexity with a agent. 04:49.720 --> 04:53.120 But there are two ways left how we can do this. 04:53.360 --> 05:02.680 So by the node or HTTP request, the next thing we'll cover is how to connect perplexity with its node. 05:02.960 --> 05:06.640 So first things first let's add the new node and set up the trigger. 05:06.640 --> 05:09.560 Or even just add the node which will be perplexity. 05:10.600 --> 05:19.010 Okay, we can just use perplexity message, a model, and then in general we can connect our credential. 05:19.130 --> 05:21.570 So as we've done it previously. 05:21.570 --> 05:23.930 So there is nothing really secret. 05:23.970 --> 05:25.170 It's nothing hard. 05:25.370 --> 05:27.930 But for now we can need to add the text. 05:27.930 --> 05:36.170 So let's say I will take this, um this chat message node, I will connect it and then um, here, let 05:36.170 --> 05:39.370 me actually move it around right there. 05:39.690 --> 05:40.410 Um, just yeah. 05:40.450 --> 05:46.090 Move it here and provide my chat input so I can provide the chat input. 05:46.090 --> 05:48.090 And then whenever I type any message. 05:48.090 --> 05:59.330 So, um, give me five marketing tools and 20, 25 then, um, you know, it processed with perplexity 05:59.330 --> 06:01.330 and it gives us the response. 06:01.730 --> 06:08.930 Um, so actually, by this way, we actually not using um, ChatGPT, but only perplexity to give us 06:08.930 --> 06:09.890 a response. 06:09.890 --> 06:16.230 So it's less, I can say, powerful, um, solution rather than creating a agent because here, um, 06:16.230 --> 06:21.670 yeah, we've got the response, which is actually, you know, not that, um, not that captivating. 06:21.670 --> 06:23.670 We've got different, um, data. 06:23.670 --> 06:26.470 So here, um, we just get the content. 06:26.950 --> 06:30.750 Also, remember, when it comes to perplexity, you've got different models to use. 06:31.110 --> 06:37.310 Um, the models I recommend is Sonar and Sonar Pro, so when you just use it, um, actually for me 06:37.310 --> 06:38.030 it's better. 06:38.070 --> 06:41.110 However, still, um, feel free to experiment. 06:41.350 --> 06:47.750 And the last thing for now for this video will be connecting our perplexity via HTTP request. 06:48.110 --> 06:51.990 So for now, um, even let me take the sticky notes so it looks better. 06:53.150 --> 06:56.070 So we have perplexity a perplexity node. 06:56.070 --> 07:00.270 And now it's time for, um, our HTTP request. 07:00.510 --> 07:01.270 Let's click here. 07:01.310 --> 07:04.230 Add the node pick HTTP request. 07:04.430 --> 07:10.670 So like each method has its own advantages and disadvantages. 07:11.150 --> 07:16.230 Um, for example, with the A agent we are able just to connect perplexity with other things. 07:16.230 --> 07:22.080 So you know our a agent that can use multiple different tools like the memory option ChatGPT. 07:22.120 --> 07:23.600 You know like a lot of that. 07:23.840 --> 07:30.000 Um, the note that actually, you know, um, there we can pick different models and also here, um, 07:30.000 --> 07:33.600 so I can say like the note is better in different circumstances. 07:33.600 --> 07:39.280 Let's say for an example, you've got your external, um, ChatGPT or previous node that is an agent 07:39.280 --> 07:44.160 or so, then you can put the content inside perplexity node and it works great. 07:44.160 --> 07:47.320 So I can say again it depends from the circumstances. 07:47.800 --> 07:55.960 Um, HTTP requests I can say is the method that we are using like an end users, um, when um, like 07:56.000 --> 07:59.880 the node perplexity node, um, wasn't available for us. 08:00.000 --> 08:04.760 So remember you can use different actually applications that even don't have them. 08:04.880 --> 08:07.640 Um, internal connection inside Nadine. 08:07.880 --> 08:09.960 Uh, for that we've got HTTP request. 08:09.960 --> 08:15.480 And in this course we are setting up a lot of HTTP requests, um, because it's very important to know 08:15.480 --> 08:16.400 how it works. 08:16.400 --> 08:23.050 But let's move on and choose the method which is post, and I prepared for you the file with all of 08:23.050 --> 08:24.170 the information. 08:24.170 --> 08:31.130 So here, um, actually you can see you've got the URL, you can just copy it or just simply always 08:31.170 --> 08:38.250 I will I will just recommend you to go ahead to perplexity documentation, perplexity API documentation. 08:39.170 --> 08:41.130 And here just check the data. 08:41.170 --> 08:43.650 You know, so their perplexity home. 08:44.010 --> 08:49.690 Um just click on API reference and you've got all of the information actually what you need to exactly 08:49.730 --> 08:50.290 set up. 08:50.690 --> 08:53.250 Um, what's the authorization, how to do this. 08:53.690 --> 08:57.330 Um, so for the simplicity, I put everything in this file. 08:57.650 --> 09:04.330 However, if you want to learn sometimes, you know, like the things how to do some specific notes, 09:04.330 --> 09:08.370 how to use them, uh, just play around with the documentation. 09:08.370 --> 09:09.330 It's really useful. 09:09.330 --> 09:14.250 For an example, we'll be using JSON and we just take this data from here. 09:14.650 --> 09:15.970 Um, it's pretty easy. 09:16.090 --> 09:21.090 Um, I think the documentation and most APIs, um, is pretty comprehensive. 09:21.090 --> 09:23.460 So again, feel free to use it. 09:24.100 --> 09:25.700 But for now, what we can do. 09:26.180 --> 09:28.460 Let's pick the authorization authentication. 09:28.460 --> 09:29.100 I'm sorry. 09:29.620 --> 09:32.740 Here we had we have actually no. 09:33.380 --> 09:34.620 We click on none. 09:34.660 --> 09:38.220 We send headers and there we provide authorization. 09:39.180 --> 09:41.540 Um, we type bearer. 09:42.220 --> 09:44.260 And by the way, you've got everything explained here. 09:44.260 --> 09:46.540 You can download it in the resources of this material. 09:46.980 --> 09:53.620 So with the bearer you need to provide your API key and move around right there. 09:54.100 --> 09:57.500 And again just paste API key API key here. 09:57.820 --> 10:04.940 Send the body and pick JSON using JSON and just paste the code I provided for you. 10:05.580 --> 10:08.700 So inside the JSON you've got it. 10:09.540 --> 10:10.700 You can put it here. 10:11.100 --> 10:14.460 And remember um, we've got the expression let me show you. 10:14.500 --> 10:21.580 Remember just instead of like we've got the system message and the user message, the user message is 10:21.580 --> 10:25.430 your prompt and the system message is kind of the system instructions. 10:25.470 --> 10:25.910 Okay. 10:26.230 --> 10:30.430 And remember inside there in the system message you can put anything you want. 10:30.470 --> 10:34.390 So as a default it's it's be precise and concise. 10:34.830 --> 10:40.110 And when it comes to the user message we've got how many stars are there in our galaxy to make it. 10:40.270 --> 10:42.030 Um actually as a variable. 10:42.390 --> 10:43.510 Um, let me show you. 10:44.150 --> 10:45.310 You can pick the chart. 10:45.750 --> 10:53.670 So there, just pick the chart, connect it with your HTTP request and inside, uh, your content, 10:54.030 --> 10:57.190 just put your variable instead of this text. 10:57.350 --> 10:59.390 So just put it here. 10:59.390 --> 11:01.270 So we've got the chat input. 11:01.310 --> 11:01.870 All right. 11:02.310 --> 11:04.990 So then when you do this you can execute the step. 11:05.670 --> 11:09.510 And it should run your HTTP request and give you the response. 11:09.510 --> 11:10.350 So you've got the data. 11:10.350 --> 11:11.510 You've got the content. 11:11.870 --> 11:14.870 Um even you've got you know you've got the search results. 11:14.870 --> 11:22.750 So the exact sources and by this way you connect perplexity insight and then by three different ways. 11:23.110 --> 11:24.230 I hope you enjoyed it. 11:24.230 --> 11:27.190 Thank you for watching and I will see you in the next video.