FREE Open Source Deep Research VS OpenAI Deep Research VS WebUI Deep Research!🤖 (Save $200/mo)

As many of you guys know I am currently on chat gbt Pro Plan which means I have access to open AI operator 01 Pro and the new kit on the Block that everyone's been talking about deep research but all this is hidden behind a $200 a month pay wall and quite frankly not everyone can afford that nor do some people want to give this money up to a close Source company but at the same time you've probably seen the hype online

And deep search is definitely great and very impressive to say the least it can spend 5 to 10 minutes and compile PhD level reasoning for different topics and it is just insane the results that some people have been able to get with it now if you've been following this channel the last few days you'll know that we actually uploaded three different videos showing you three different ways that you can get around that $200 a month and use open-source alternatives so we've

Covered so far open-source deep research a very powerful CLI tool we've covered web UI which has a deep seek functionality as well as an operator alternative to open a eyes operator and then yesterday we covered another wrapper which is a gooey interface for the CLI open deep research project now I've been getting a lot of different comments because many people are interested in this deep research and people are wondering what is the best tool to use for deep research should you

Go with the open- source tools should you pay the $200 a month well in today's video I'm going to show you exactly how to set up each and every one of these different tools and I'm also going to do a comparison between each and every one so you can decide which one is best for you and your particular use case and if you should go with the open source versions or the closed Source versions I'm going to compare all these tools

So let's dive in and start doing some deep research now all right guys so all links I covered in today's video will be in the description down below I'm also going to be using this document here that I made and we're going to be putting all the different links here the resources the prompts as well as the actual research output from today's video so you can analyze the comparison on your own time if you want and see which tool gets the

Best output now if you want access to this entire document it is going to be 100% for free I'm going to be posting it in the stride AI Academy if you're not familiar with that that is our free community on this channel where you can network with other like-minded AI entrepreneurs or AI developers in this space and I'm also posting all the templates resources tools Frameworks videos and behind the- scenes stuff on this channel in this community so you're definitely going

To want to join just go to the link that I leave down below for the stride community and request to join and then you'll be able to access the document right away all right guys so before we actually dive into the test I want to quickly go over setting up these tools now I've linked all the separate videos that I've done to each one of these tools in depth so if you want to see a step by step walkthrough on setting

Up each and every one check out the full video for the specific tool I've also linked the githubschool see the firec API docs here or browser use so let's talk about the actual open- Source deep research CLI version right here all right so here is the GitHub repo this is an AI powered research assistant that performs iterative deep research on any topic by combining search engines web scraping and large language models this is how it specifically works you have input you

Have a depth parameter a breadth parameter and a user query it starts doing deep research gathers Ser queries pulls different results uh has different directions and or learning uh analyzes the depth and then either performs a markdown report or next directions has prior goals new questions and learnings all right so this tool is great but it does just use the CLI and here's the setup here you simply clone the repository you install the dependencies running mpm install you set your environment

Variables in the env. loal file make sure it's env. loal so you're going to copy the example one and then paste in your uh API keys and then you're simply just going to run npm start to actually run the assistant and then it's actually going to run within your command line I went through that very quick you know I like I said if you want to use a CLI version check out my video on it and the instructions are right here

In the GitHub repo for you but I know most of you would prefer a nice gooey interface which is why we're going to be covering the another wrapper project where if you didn't see my video yesterday it's essentially built with this CLI in the back end to do all the Deep research but it's using their nice user interface gooey and this is what it actually looks like so you can go to their Cloud version right here another rapper.com and you'll be

Able to actually access um and do deep research if you just configure your API keys right here within their Cloud version and I'm going to show you this in just a second but I'm going to show you on the actual self-hosted version well you can uh change your breadth and depth parameters right here so you have it's cool because you can't do that functionality with an open AI deep research it just actually goes about and does the specific deep research for

You know could be 5 minutes or could be 20 minutes depending on the specific question you ask it but here you can actually change these parameters which will affect how indepth uh it's going to go about doing your research which is very nice so here's the GitHub repo for that guey right here and I'm going to quickly show you how you can actually set it up and self-host it before we run some tests on it so you're simply going to do

The exact same thing I just went over we're going to get clone this repo right here then we're going to change directory into open deep research and then we're going to run npm install so you can literally just copy all of this if you don't know too much about commands paste it in right here and then run it I'm not going to do that because I've actually already ran it but that's what you're going to do once you do that you're

Going to want to configure your environment variables so I always like opening it either in cursor or in uh VSS code so you're going to run code dot or cursor dot okay and then you'll see a.v. example you can make a copy of that and rename it to EnV and then you're going to paste in your opening AI API key as well as your fir craw API key now if you don't have a fir craw API key simply go to fir

Craw right here sign up get a API key and paste that bad boy in now the one downside personally is fire craw is great but there is some limitations and there are credits so if you're using this a lot you may actually run out of credits or have to actually pay for credits so I personally usually prefer using something like crawl for AI and now at the moment there's no native integration but I actually may look into playing around with some

Of this and seeing if I can maybe tweak some of this and for for you guys and then do a video on that so if you want me to do that and see if I can get something ready for you guys maybe let me know in the comments down below now once you do that you're simply just going to run npm runev so it's the exact same as the you know CLI version and it's going to start on Local Host 3000

Okay and boom now we have our Local Host version right here we're going to come back to this in just a second when we start running our tests all right so the last open source project right now is web UI open source and guys if you know of any other open source deep research projects or whatnot let me know in the comments down below and I may do a video on it so um let's go ahead and check out this repo

Here all right so this project builds upon the foundation of browser use I've done a video on browser use it's essentially like an open AI operator and you can build out custom code within python to use browser use it's really awesome really cool project and it's made by the same people who made that but web UI is a nice gooey for that and it also has deep research capabilities which is really really cool this is a very powerful tool so we're

Going to start by cloning the repository and then changing directory into web UI so exact same process that we always do all right next you're going to want to have UV installed so this is for managing your python environment so simply just go here make sure you have UV installed and then you're going to run in the command line UV and then VV for virtual environment and then python 3.11 make sure you have python installed too if you don't and then

Depending on what system you're on whether it's Linux or Windows or Mac you're going to run one of the fulling commands so on Windows you're either going to run this for the command prompt or this for Powershell right here um to activate your virtual environment and then for macro Linux you're just going to run this Command right here okay all this is literally Linked In the description within the repo and if you're having trouble with any of these installations check out

My individual video on that specific project so you can actually get some help with that and then you're simply going to install the dependencies so install python requirements like so run this command and then you're going to install playright this is all within the same terminal and then you're going to configure your environment variables just like we did so I like opening it in cursor or vs code so once you have that open you're going to see e.v. example you're going

To make a clone of that and you can add your open AI API key your anthropic API key or Google API key you have a lot more options than some of the other Alternatives which makes me really like this because you can really use this with any model almost well really any model cuz you have an own llama endpoint right here so you can use it with open source models as well and and you can also change your open AI endpoint

So you could Point any provider you want and actually use an open AI operator or open AI deep research alternative for free which is very very nice and Powerful then once you got those API Keys set up you're simply just going to run this right here so python web ui. and then the IP right here and then the port right here so you can simply just copy this and run it once you run that Command right here it's going to start

Everything up on Port 7788 Okay and like I mention I'm leaving the docs in the description for browser use so if you need any help you can check out the docs right here it explains every single thing all right so now we have all the different ones that we're going to be testing setup we have browser use web UI right here which is really nice we'll go over this in just a second we have another wrapper which is built upon the

Open deep research CLI project and then we have open ai's very own deep research the $200 a month plan right here which we're going to be testing and seeing the comparisons of all three because of course A lot of these different ones have from bells and whistles but when it really comes down to what matters it is the output and the quality of that all right so we're going to start with opening eyes deep research so you'll see here the prompt

Is I'm doing a YouTube video comparing the Deep research of three different tools the first tool is opening eyes new deep research tool $200 per month but you also get access to on1 Pro higher sore limits and probably more openi deep research and then giving the blog post link the second and third are open source alternatives to opening eyes deep research the second one is web UI by brows browser use so web UI by browser use right here so a link

To the repo a link to browser use website right here and mentioning that you can use nlm with this project the third one is open source deep research right here which is a CLI tool and then linking to the GitHub repo right here it uses fir crawl linking to the fir crawl API docs right here but someone made it gooey for this project so you don't have to use the CLI and it's called another rapper right here open source deep research

And linking to the repo and they even have a cloud version which is here and then linking to the cloud version please do it in-depth comparison between the closed Source open AI deep research and the two open source Alternatives and weigh the pros and cons as to which one my viewers should choose now when you're using deep research you have the option to select your models so typically you're probably going to want to use either 01 Pro or 03 mini High

For This example we're going to be using 03 mini high now the reason we're using this is because one it is super super super powerful it's a lot faster than 01 Pro and this is also going to be the same model that we're going to be using to test the other deep research tools so we're giving it an even playing field between the comparison of the output so we'll go ahead and run this right here and you can see that we

Get a few different followup questions so it's saying that's a great idea to make sure we can do a useful and end up comparison could you please clarify a few different things here okay so I'm going to use these uh followup questions for pretty much all them so we can give every single project the exact context for their search so to go over these follow-up questions we have evaluation criteria so are there any specific factors you'd like to focus on accuracy

Speed easy use integration options cost flexibility data privacy I said all the above target audience is your target audience more technical devs researchers or general users I said both so have both options but more devs and researchers probably and then use cases do you want to focus specific use cases such as academic research market analysis content creation or something else we're going to go all the above and then depth of comparison would you like a simple pros and cons breakdown or

A detailed structure analysis with performance benchmarks screenshots and potential user testimonials I said detail is possible but also have a simple version and then preferred format would you like a summary table in addition to the written comparison I said yes and also give unbiased opinions based on what each specific user may want to use for their specific use case to get the best output okay now we're going to go ahead and click Send here and as you can see it's finally

Started its deep research right here so as this is going let's go ahead and start doing the same thing for the other ones so right now I'm on the another wrapper one which is using the open deep research CLI on the back end and we're going to paste in that same prompt right here and we're using 03 mini you could also use some other models right here unfortunately this one is limited it's you're not going to be able to use like

O Lama or anything like that out of the box you could probably tweak it but unfortunately you can't use it out of the box now the one cool thing with this is we can adjust the breadth and the depth for this search so by default it's on four bread and two depth I'm going to go ahead just for the sake of this test and do six bread and three depth but you know you can make this even more in depth or

Less in- depth depending on these settings right here now we're going to go ahead and send this okay so it's asking some different followup questions I'm actually just going to go ahead and paste in the same followup questions we Ed for the open AI one just cuz I want to give it the exact same context and now we're going to go ahead and press send and now it's going to say starting in-depth research based on your inputs so the cool thing

With all these like if we're looking here in the open AI one you can see all the different activities going on right here on the right hand side so you can see it's going to W combinator YouTube GitHub of course and you can actually go through and track some of the activity and see how it's doing its reasoning as well as it's researching and you can see all the different sources right here that it's compiling which is really cool and you

Can do the same thing pretty much with the open source versions too so we can see it's doing uh different search queries right here and then doing different research finding different results processing them and then generating new learnings and then it's going to take those learnings and then be able to incorporate that into its new research that it's doing right here so as this one is going let's go ahead to our final one which is browser uses web UI version now

I'm not going to go over all the different settings here as you can see there's a lot of different configurations you can use you can do agent types here maximum run steps so just like how you can change different configurations with the one we just s for open source you can actually do the same thing with browser use which is really nice and as you can see here with llm configuration we can select any llm we want we're going to use

Open AI right here and we're going to use 03 mini just like we use with all of our other ones but like I said you could use AMA different ones like that and I'm going to show you actually in future videos how to use this with other models and showing you the different outputs that you can get using deep research with maybe some open source models or with some other closed Source models and I really like this because we can actually

Tweak a lot of different things like the temperature um and all that good stuff over here in browser settings you have some different options right here so we can use your existing browser keep browser tabs open run browser without guey so in headless mode um you'll see once we actually start running this it's actually going to open a browser if we're not running it in headless mode and you're going to actually be able to see the agent do deep research within

Different Chrome browsers so that's kind of cool um and you can actually kind of see like it's like a real person doing research and you'll see the queries and everything so I really like this you can see the recording path Trace path agent history this has a lot of settings so if you're going to want to use this one I'd recommend checking my individual video I made on this where I go over all these different settings here you can run different

Tasks right here if you're just using like open ai's operator alternative so this will actually do tasks we're going to use the Deep research to do research so it's going to use that AI agent to open the browser and then take that information it's going to scrape the the page and then be able to do different queries and then generate the report that way here you see all the results information the recordings will show up here and then if you want

To import or export a configuration you can do that here so I'm going to go ahead and paste in this same exact prompt right here all right so we're going to have the max search iteration to three and the max query per iteration to two and we're going to go ahead and run deep research right now and as you can see this one's a little bit different when we actually run this it's opening up two browsers on our desktop okay and

This is actually where it's going to do the queries so we can see open AI deep research verse web UI versus open source deep research features comparison and we can see the same thing right here with a different query in another browser if we look in our terminal right here this is where you'll actually be able to see the live logs of what's going on so new memories task progressions future plans summary different actions etc etc all right so while we

Are setting up web UI open deep research as well as open AI deep research are actually complete so the one thing with this another raer right here um you'll be able to see the logs right here it isn't as in-depth as open AI logs you can also go to the CLI right here and you'll see the logs that are how the CLI tool works and you can see it's doing different searches having different research routes Etc um the one thing is

It doesn't show the time from what I can see within the guey which is kind of not the best I wish it showed how much time it did for its research but if you go ahead to the uh CLI you'll be able to see that it did a request right right here for this time right here which is 15632 seconds which is equivalent to 2 minutes and 3632 seconds now if we go ahead and look at opening eyes deep research this

One actually went for 11 minutes now like I showed you before you can change the breadth and depth to make it you know a little bit more in depth or a little bit less in depth whereas open AI you can't control that now with another wrapper right here if you scroll to the bottom well one thing you'll be able to see the whole entire deep research right here which is really nice in the UI but you can also just download the

Report right here so it we'll download a.txt file if we go ahead and open this up you'll see the whole entire research report in markdown file which is really nice you could feed this to another llm if you want or do whatever you want with this actual data now I went ahead and just paste it in the research report right here so we'll go over it briefly I won't be able to go over every single thing but you can see here

In-depth comparison between deep research tools you can see a table of contents right here so we we have our introduction um you know giving an intro to all three tools right here an overview of the tools so open AI deep research browser use web UI so it's going over things like the model the cost the features the strengths limitations we have browser use webui so the model the cost the features strengths limitations open source deep research CLI and GUI going over

The same things right here and then evaluation criteria so accuracy Speed and Performance ease of use use integration options right here cost and cost benefit analysis flexibility and customization data privacy and security and then deployment models and use cases so deployment modalities uh use cases right here and then a summary table so we can see here a full breakdown on a bird's eye view with cost accuracy speed ease of use integration flexibility data privacy and then recommendations so here with the

Another wrapper open deep research for General users it actually recommends open AI deep research and then for developers and researchers it recommends the open source Alternatives and I don't know if I had 100% agree because um you know General users may not want to pay $200 a month all right that's a huge factor for someone that's generally using it a researcher uh or a developer may actually want to pay that money because it's an investment into their most likely their job

Or their business all right so pretty in-depth right here as you can see it uses a bunch of different sources here so it actually pulled from 19 different sources here now if we take a look at open ai's deep research it actually pulled from 27 different sources here and it's very very long and intensive so let's go ahead and paste this into our dock also one thing to note the open source deep research another rapper was 2,346 words and 18,957 characters

And the open AI deep research report was 16,336 words and 100,313 64 characters so it is a huge report so we can see here the alternative comparison a summary comparison table right here that's going over accuracy aspect speed it's going in depth like it's pulling out different Benchmark tests right here and getting a lot of different data you can see the speed the ease of use so very easy moderate moderate um giving the direct links right here which is nice integration

We can see a cost and for each specific one it's going very in depth for each specific point flexibility right here data privacy accuracy right here going over different links and sources and I like how it gives the sources directly in the report instead of just giving them only at the end and then speed here so more different sources very very indepth like I couldn't even go through this all of course ease of use okay so integration options cost each section

Is super huge too like all these are very long flexibility data privacy and security and then pros and cons breakdown so open AI close Source $200 a month Pros no setup required and easy to use integrated with chat GPT high quality models thorough and reliable research process citations and sources provided maintenance and support FS very high costs limited usage slow report response time no customization or flexibility closed ecosystem data privacy concern required internet and public data a one siiz fits-all model

And then browser use right here open source free and open source multi llm support runs locally browser automation capabilities parallel processing and efficiency userfriendly interface highly extensible integration flexibility no fixed limits and then cons requires technical setup resource requirements maintenance overhead not as polished as commercial product learning curve for full utilization security consideration dependent on external apis potential to misuse if not careful and then the open deep research CLI with gooey rapper Pros free and open source replicates opening eyes agent

Logic adjustable parameters modern guey parallel and efficient search citations and markdown output self-host or privacy fast adaption and Community Support flexibility to integrate or modify lower resource overhead and then cons initial setup needed requires API Keys limited built-in search choices potential rate limits and cost surprises maturity and stability less General than browser use monitoring and intervention dual dependencies open Ai and Fir crawl and then documentation support is community based then going over use cases based on recommendations and then a summary

Table of key differences and that's it for that one that's very very long just as we finish that off the web UI report is done so we can go ahead and see this now this one is much much smaller than the other one so if you want you can actually download the report right here but you'll also be able to access this in the folders right here so if we go to our deep research folder right here in our temp folder

You'll see this is our last query and we can see the different query results these are all the different scrapes that it got for different sources and then you can see the final report here you can also see the record info right here so here we actually got 1 2 3 four five six we only got six sources I will say guys I probably should have put it to higher different search iterations and Max queries I actually did do that initially

But I was getting an error are you tired of pouring thousands of dollars into appointment Setters only to watch leads slip away imagine having a team of elite sales agents booking qualified appointments for you around the clock no more wasted time on training no more frustration with performance and no more draining your budget on inconsistent and expensive call centers introducing stride agents AI powered appointment Setters that work 24/7 never get tired and book appointments while you sleep trained on thousands of

Successful conversations our AI agents El perform human teams at just one tenth of the cost join the ranks of businesses that doubled their appointments and booking rates in just a matter of weeks don't get left behind in the AI Revolution visit stride agents.com now and transform your entire sales process with Cutting Edge AI technology it's time to accelerate your stride with AI agents so see I was getting these failed uh tries right here constantly so I just put it back to

A little bit lower and then it was working so I don't know if that's just something with me or maybe it's some issue that it has okay now unfortunately at least as to what I can see I don't see a spot where you can see the specific time that it was actually run for I know when it was running it did show but now that it is run I cannot find it anywhere and I don't see it in the logs either

I did check and if you did have time stamps on you can maybe see it but I will say that it did take quite a long time it took maybe I would just guess maybe about 5 minutes at least and it only generated 661 words and 4,967 characters so we can see here comparative analysis overview of tools uh comparative features so cost and flexibility usability and setup customization and control pros and cons and conclusion now also too I will say that

This is the only one that didn't ask me a follow-up question once I gave it that initial prompt so if I was to redo this again I would probably give that prompt with the followup questions so to make sure that you give all the contexts out front you know think of the different scenarios um because it won't ask you a followup question unfortunately so all in all just to run over some of what I seen with doing these tests webui does

Give a bit of a shorter answer I found I mean you I did have the lower settings on so maybe giving it more settings it's going to give a longer answer of course but me personally within this test I say that it probably got the maybe the worst output but maybe if I did this test again and gave it more of that initial context that I gave in the follow questions maybe it'll give a better output it did get the least

Sources but keep in mind it does have a lot of good flexibility with being able to use different models and test it so I'm going to keep using it I think it's still really good now open deep research did a really good job with its report it's it's in-depth I could have made it longer if I gave more breadth and depth but it did a really good comparison to say the least and it cited you know a bunch of different sources

19 different sources which is really good and then honestly I would have to give the win in terms of the output here to opening eyes deep research search I mean it's very very indepth and I think it probably maybe even like too in depth which is good but you could also you know tone down some of the prompt if you want to get less information and honestly too guys if you want to really get in-depth nitty-gritty comparison you can go to

This Dock and read some of these comparisons that these AI models generate especially the open AI one it is pretty damn good I think like I said guys if you want access to this complete document just join our free stri a Academy I'll leave a link down below where you can access it other than that guys that's pretty much it for this video let me know what your thoughts are in the comments down below are you going to be buying open

AI deep research for 200 bucks a month or are you going to use some of these open source Alternatives I'm personally going to be using both I like open source I really love open source um one there's no limits on these you know open AI deep research only gives you 100 queries a month yes it is really really powerful and if you do have the funds and you're someone that's serious about AI uh doing research and whatnot it may definitely be

Worth the investment me personally I do not regret investing the $200 a month to use this tool as well as the other tools right now it is pretty damn impressive but I also really like these open source Alternatives and I'm going to be doing different tweaks with some of them testing out different ways to use them with different LMS and whatnot in future videos so make sure to stay tuned for that let me know what your favorite one is in the

Description down below and guys if you just need a basic deep research agent or something that is really good but maybe you just don't need to spend $200 a month then these open source alternatives are very very powerful nonetheless other than that guys that's pretty much it for this video on this in-depth comparison let me know what your thoughts are in the comments down below if you're new to the channel we upload videos all the time on ai ai agents AI

Coding business growth Marketing sales so if you like that type of content you got some value here I really appreciate it if you subscribe like the video and comment down below thank you guys so much for the recent 17,000 subscribers 20K on the way guys and then like I mentioned guys if you haven't already joined our free Facebook group and Discord Channel stri community.com I'll leave a link down below and then definitely check out our stride AI Academy where you can

Get all these free resources and a bunch of different stuff in here and then also too guys if you run a business and you need custom AI Solutions or custom AI agents like AI appointment sets Ai call centers or whatever the case may be book a call down below at executive.com apply and we can see if it's a fit or not or if you're someone that's looking to sell AI Solutions AI agents AI services to other business owners and you want

Our blueprint and protocol on how to actually do that book a call down below and we can see if it's a fit to work together other than that guys I will see you in the next video keep hustling keep grinding and of course guys accelerate your stride take care