FREE GPT Researcher AI Agent Powerhouse!🤖Open Source Perplexity Self Hosted - Perplica, Morphic
In this video
- 00:00Introduction: Free Open Source GPT Researcher AI Agent
- 00:29Previous Videos on Perplexity Alternatives
- 01:07Overview of GPT Researcher Features
- 01:49Autonomous Research Agent: Key Differentiators
- 02:22GPT Researcher's Architecture and Research Process
- 03:17Installing GPT Researcher: Python, API Keys, and Setup
- 04:37Customizing with Various LLMs and Search Engines
- 05:25Virtual Environment Setup and Installation Commands
- 06:13Demonstration: Researching Elon Musk's Robots
- 06:48Customization Options: Report Length, Tone, and Sources
- 07:14Ad: Stride Agents AI-powered Appointment Setters
- 07:42Research Options: Web, Documents, or Hybrid
- 08:41Live Demo: GPT Researcher in Action
- 09:25Report Generation: Markdown, PDF, and Docs
- 10:05Second Demo: Researching OpenAI's Swarm Framework
- 10:36Advanced Customization: LLMs, Docker, Langchain, and Vector Stores
- 11:08Final Thoughts on GPT Researcher vs. Perplexity
- 11:57Outro: Daily AI Content, Community Links, and Stride Agents Promo
Links from the video
- docs.gptr.dev/welcomedocs.gptr.dev/docs/welcome
- gptr.devgptr.dev/
- app.tavily.comapp.tavily.com/
- arxiv.org/2005.11401arxiv.org/pdf/2005.11401
- arxiv.org/2305.04091arxiv.org/pdf/2305.04091
- FREE: Morphic AI - Perplexity AI Alternati... %20morphic.sh/
- console.upstash.comconsole.upstash.com/
- miurla/morphicgithub.com/miurla/morphic
- serper.devserper.dev/
- console.upstash.com/vectorconsole.upstash.com/vector
Transcript
2,299 wordsWhat's going on everybody Josh pook here and in today's video we are going to be looking at an AI researcher agent this is a free open- source alternative to perplexity it is called GPT researcher say goodbye to hours of research say hello to GPT researcher I'm going to show you how you can self-host it let's dive right into [Music] it all right guys so some of you guys may have seen some of my other videos and I've done videos on a
Lot of different perplexity Alternatives so I've done videos on uh open perplex I've done videos on local search I've done videos on morphic I've done videos on llm answer engine I've done videos on perlex we've covered a lot of different perplexity Alternatives and I'll leave links to these videos as well down below so if you are looking for open source gxy Alternatives you can check some of those out but in today's video we're going to be looking at GPT researcher and
Here is their website uh you can check it out all links will be linked down below they claim to be the number one open- source AI research agent elevate your application capabilities with the leading autonomous specialized research agent built for multi-agent Frameworks delivering real-time accurate and factual results at speed over 1 million downloads and hundreds of contributions worldwide for support for all llms use any search engine research local documents generate long reports export to any format multi-agent support and one of
The reasons I decided to do an video on this tool is because if we look at the example here you'll see that it's querying the gpg researcher agent and it's pulling a bunch of different sources and deploying a few different agents and it looks like it's doing some pretty extensive research and then it's going to provide the output in very quickly that's what kind of stood out to me with this tool opposed from some of the other open source researching uh
Perplex the Alternatives we covered on here this seems more of like a researching Powerhouse that's the angle that they're going for all right so GPT researcher is an autonomous agent designed for comprehensive online research on a variety of tasks the agent can produce detailed factual and unbiased research reports with customization options for focusing on relevant resources and outlines so was inspired by the recent plan and solve and rag papers which I'll leave links to down below so you can check these
Out if you'd like GPT researcher addresses issues of misinformation speed determinism and reliability offering a more stable performance and increased speed to paralyzed agent work opposed to synchronous operations so the architecture the main idea is to run planner and execution agents whereas the planner generates questions to research and the execution agent seeks the most related information based on each generated research questions so similar to perplexity finally the planner agent filters and Aggregates all related information and creates a research report a
Few features generates research outlines resources and lessons reports with local documents and web sources and generate long and detailed research reports over 2K words Aggregates over 20 web sources per research to form objective and factual conclusions includes both lightweight HTML CSS JS and production ready nextjs plus tailin uxui okay so let's get into the actual install of this so there's a couple different ways to install you could use Docker you could use pip or just going to do the uh basic
Install here so you're going to need python 3.11 or later so you can go here and download it if needed and then you're going to download the project and navigate to its directory so you're going to get clone the repo right here and then CD into GPT researcher so I actually already did that so once you do this you can set up your environment variables so you can either do code dot to open it up in vs code like so and
You'll see um there will be a env. example and you can just either um change the name or copy it then change the name and then put your uh open AI API key here and you're also going to need a Tav API key okay or you could just export it to uh run these commands to export it to your environment variables in your OS either way works now tavali search API is a search engine specifically optimized for llms and rag aimed
At efficient quick and persistent search results now it says here the default llm is GPT for open AI but you can use any llama or any model pretty much such as Claude o llama Gemini I mistro you can visit the L documentation right here I'll leave a link to the docs down below and same thing for the search retriever the default is Tav but you can use things like Duck ducko Google Bing search API serper search X archive even right so
Pretty cool stuff um you can visit the retriever documentation right here um Tav is good I believe they give you a bunch uh a decent amount of free credits to start off so you can pretty much use it for free if you want to test it out or if you want to try of the other ones you can do so okay once you once you do that you're going to want to uh install the requirements so for me I did so
In a virtual environment so I just set up a virtual environment right here you can do python DM and then venv and then I just named it my EnV and then I just activated it right here which is this script if you're on Windows if you need help searching setting up a virtual environment just ask chat gbt you could use this um this way right here you could also use cond of course and then once you do that you're just going
To run pip install D r- requirements.txt to install all the different requirements right here and then once that's done you're going to run this command python DM uvi corn Main and then colon app-- reload right here this is going to start the fast API and it's going to start the agent on Local Host Port 8000 so now I'm on Local Host 8000 and you can see here we got it set set up all right so I could go ahead and ask
It a query right here so I'm going to ask it uh to research about Elon musk's new robots launched in October 2024 which was launched yesterday um and then what type of report would you like to uh like me to generate so we have different options here we could do summary which is short and fast um 2 minutes and then detailed which is in dep then longer 5 minutes and then research report so just for you know to test this out
To start we can do short and fast and then we can change the tone so we could do informative providing clear and comprehensive information on topic persuasive we could have a bunch of different options here which is pretty nice that it gives you all these different options a lot of the other open- source perplexity tools don't give you options like this so this is pretty useful 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 outperform human teams at just oneth 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 agent now and transform your entire sales process with cuttingedge AI technology it's time to accelerate your stride with AI agents and then what sources would you like me to research from so you can use the web um you can also add documents so please make sure to add the dock path in the environment variable pointing to the documents folder so this is
Pretty cool as well that you could actually add your own documents and set up like your own rag so you could have your own documents and also do the web so you could do the just the web which is kind of like a perplexity search you could do just your documents which is pretty much like somewhat of a Rags type search um and then or you could do a hybrid approach so you could probably uh get some pretty interesting results if
You maybe were doing a hybrid approach you give it some basic context within your documents some personal uh stuff that is customed to whatever you're researching on whether it's something related to school or a coding project your business and then you also are asking at stuff that it would uh want to search on the web and you can probably get some pretty custom stuff there so and then you just click on Research so we're just going to do the web for
Now if I click research we can see thinking about research questions to ask then starting the research task and then using the technology agent right here okay it's saying it's going to conduct the research now it's conducting different research right here all right getting different sources getting relevant information okay now it's actually building out the research report here so we can see Elon mus new robots launched in October Robo van Optimus humanoid robots challenges and skepticism conclusion references so we can
See all the references here now cool thing is too we can copy the clipboard in markdown we can download as markdown or we can download as PDF or docks which is pretty cool all right so now we can do a different one here such as open ai's new AI AG framework swarm and I'll probably do some videos on this too in in very soon um because this is pretty interesting if you don't know about that they just dropped that and pretty
It's open source framework for AI agents and AI swarms so um we're going to do a detailed in-depth and longer um report this time and all the uh other stuff is going to be the same we're just going to research now okay so we can see it's generating different queries here searching different sources all right getting subtopics here writing introduction potential applications future prospects and developments all right so when you do the long one as you'll see right here it can
Actually get pretty long which is kind of cool um but it's also like going to use more tokens so and like I said I'll leave the docks down below because you can do you know customize this in different ways you can set it up with Docker you can customize with Lang Chang in Vector stores you can change the llm provider you can use AMA run everything Loc change the retrievers multi-step agent Frameworks a bunch of different stuff so I think this
A pretty cool project they have over 14,000 stars on GitHub so um pretty interesting pretty cool all in all I think it has its strengths and weaknesses just like all the other open source perplexity Alternatives do I think it beats perplexity I mean I'd probably prefer perplexity over this but I think this is definitely it has its own use case and it kind of like that Powerhouse researcher and it's great because it's open source too you can self host it can
Change different models um other than that I think it is really good in terms of that but when it comes to doing the research it actually does a pretty good job all right so that's pretty much it for this video guys if you're new to the channel we upload videos every single day on AI automation business growth Marketing sales Etc if you like that type of content then make sure to drop a like And subscribe to the channel to stay up
To date with the daily uploads got a couple cool videos coming shortly guys that I've been working on so stay tuned for that should be in the next couple days or so and then if you haven't already joined the free community stri community . our free Facebook group free Discord Channel link for that will be in the description down below let me know if you guys have tried GPT researcher down below or if youve tried some of the other open source
Flexity Alternatives like perplex llm search all the other different ones I mentioned and let me know which one is your favorite or do you just prefer perplexity and then also too guys if you run a business and you have leads and you need AI appointment Setters booking appointments for you on autopilot then check out stride agents.com I'll leave a link down below other than that guys I will see you in tomorrow's video keep hustling keep grinding and of course guys accelerate
Your stride take care