FREE: Langflow AI RAG Agent GUI Drag-N-Drop Builder🤖 (Python) Open-Source FlowiseAI Alternative!🚀

What's going on everybody Josh pook here and in today's video we're going to be taking a look at langlow a visual framework for building multi-agent and rag applications it's open source python powered fully customizable and model and Vector store agnostic it's an alternative to tools like flow wise Ai and Diffy AI which we covered on this channel we're going to show you how to install it and let's Dive Right In [Music] okay guys so links for all the link uh everything

Will be in the description down below here is Lang flow's GitHub if you haven't heard of Lang flow um you can check it out you can check out their GitHub as well as all leave the docks and everything you really need to know but um this is Lang flow as you can see here and it gives you some information on usage and how to use it how to install it and I'm going to show you that in just a second you

Can also Deploy on Google Cloud platform uh on Railway on render kubernetes many different things right but um you can also just deploy it locally and uh yeah here that's pretty much it with the GitHub here's kind of what their here's their website so the easiest way to create and share conversational AI discover a simpler way to build around found foundational Models All right so we can kind of see the UI they have a really nice UI I like the feel

Of it you can see prompt variables Dynamic input are introduced using curly brackets fine tuning harness the full power of llms by easily find tuning them from spreadsheet so this is really cool um I'm actually very excited to use Lang flow um and yeah build out more stuff in the next little bit group components so higher level components naturally emerge from AI building block and share your Creations custom components go beyond the surface and code your own components not just the

UI not just the logic um Lang flow is dynamic a dynamic graph where each node is ex an executable node its modular and interactive design Fosters rapid experimentation and prototyping pushing hard on the limits of creativity and it's python based so that's the really cool thing I like about this it's python based um so yeah it's python python is the go-to language for data scientists by being python native langlow leverages the most powerful data manipulation and machine learning libraries out there

So yeah um that's one that's like one of the main reasons I'm actually really excited about this just because it's built on Python and uh yeah so you can I would like if you're going to be using this I joined their Discord you can join their Discord here check out the GitHub maybe check them out on X it also did a blog post kind of announcing um some of their new features here they got Lang flow 1.0 which is out so

They also have a cloud service so you could sign up for the cloud service check that out it's through a data Stacks so it's free to use and remains agnostic to any database data source or API so you don't actually even need to like run it through the command line pip install or python install Lang uh langlow whatever the case may be um you could actually just go sign up for data stacks and I'll leave a link down below to do

That so if you just want to do that you can do that test it out whatever or you can just run the command and um yeah I'll leave this blog post that they released in the description down below too and yeah just to quickly go over um you know we've talked about flow wise AI as well as Diffy AI I've done videos on that so you want to check those tools out check out some of my other videos on this that

And they're very kind of similar but they're different I'm not going to like do an in-depth comparison in this video and I'm still like really seeing um how they are different but um what I would say initially if you're wondering like the the thing that I really am excited for langlow about like I really just said is it's built on Python and and uh it it what they said on their website it seems like that is the case where it is

Like one prom variable is fine-tuning group components and um it's very modular to Foster rapid experimentation and prototyping and dynamic where each note is ex an executable unit so I really like that I you'll kind of see a bit more of that once I once you see the uh install of it and you see kind of how it works in the UI but yeah let's go ahead and just get right into it and install this thing um but yeah so it

Is somewhat of alternative to these tools I'm still going to use all them or you know be using different ones until I really find either maybe I like one the best or maybe I'm just going to use all of them right so they they definitely do have their own strengths and and whatnot but anyways before we get into it I'll leave links below down for this but you're going to need either uh you're going to need python installed you run these

Commands pip install right so links for that will be in the description down below if you check out their docs right here for langlow you can see it they they mentioned the data Stacks option and link for the docks will be down below too but uh all you really need to do to install this locally is run this command python DM pip install langlow dasu so do this once you obviously have python installed or um you can run uh do it

With Pip X so there's that option too now pip yeah so anyways run that command so you go into your command line and you literally just run it and it will go ahead and start installing everything which I already did once you're done that actually very simple you just go python um no then from that point let me just see uh once you're done that so you run this right here and then you do python dasm Lang flow run all right

Once you do that um it will basically show right here saying that it is you can access it at port 7860 on your Local Host so let's go here and this is what the dashboard looks like all right so kind of similar to some of the other tools that like Diffy or like similar UI um kind of like to Diffy and then similar to um to flow wise as well in some regards but um we'll quickly go over some of this

So you can kind of create new projects here you can add folders here so you can organize your projects you can see either all flows components um so let's go ahead and create a new first off I'll just cover a few things so you can go into settings here and you'll see General and change your store API key so you don't have an API key set you need to go to Lang flow here so let's go here I'm not going to

Add my API keys at the moment but um I believe if you connect it to langlow API key then I think that allows you to connect to other uh Services as well um not 100% sure but yeah we have Global variables here Google API key pinec cone API key so you can add stuff here like that and uh messages inspect edit and remove messages to explore and refine behaviors shortcuts you see shortcuts here keyboard shortcuts toggle on and uh dark mode

And light mode so of course we're going to use dark mode see your notifications up here so just to cover some of those Basics and um yeah so once we're done now that we're kind of went over that there's my collection and storees so in order for store to yeah you need the API key so let me just quickly okay so I actually just created a langlang store account so you just can create one um and then you can go back

Over here and go to your settings and paste in your API key here okay then you save it and now if I go to store we can see um bun okay so this actually yeah pulls in the different uh different templates I believe from like the community or the store or whatnot you can see some of these chat oama rag based llm SQ well database open AI Runner uh YouTube transcriber so a lot of like there's over 56 uh 36 results

So this is in beta right now let's go to collections here go a new project so you could either start by blank flow or you could select one of these templates just for you know the sake of this video we'll go to Vector store rag just because everyone loves a good rag um and I'm going to be covering more about rag I've done videos on Rags already I'm going to be doing more videos on rag graph Rags um knowledge like all

These different stuff so stay tuned for that if you're new to this channel we upload videos every day on AI automation business growth Marketing sales Etc so subscribe uh to stay tuned you know up toate with new AI tools all this stuff tutorials so and um yeah and so now that we're in the UI Builder we can see um this is our Rag and I'm not going to go through each and every step d uh exactly but uh and I'm not

Going to connect everything obviously you have to add your API keys for open Ai and you're using Astra DB depending on what database Vector store you're using you're going to have to set all that up with your credentials um for depending on yeah so like opening I key goes here API key goes here can select your embedding model as your database uh application token API endpoint then it's going to parse the data and up here is a chat input so chat

Input embedding all right embedding vectorize and then search input from the chat input parse data convert data into plain text following a specified template all right prompts create prompt template with Dynamic variables generates text using open Ai llm and then chat with uh output and then here is file to Tech splitter so we're chunking with a 200 chunk overlap and chunk size of a th000 and this is going into Astro DB vectorized and um well yeah uh same embedding right here

So pretty basic rag right here um or vector store and let's see so just to go over some of the components we have inputs such as chat input text input outputs such as chat output text output prompts right here datab uh data so API request directory file URL web hook input models we got AI ml API Amazon Bedrock anthropic Azure open AI cohere uh Google generative AI Gro hugging face API all these other ones mistl Nvidia olama um open AI ver

Vertex let me just curious go here and you would put your base URL um you can select your model right so yeah this is actually awesome uh vertex AI uh helpers so chat memory go like this um so you can do external memory messages text memory create list combine text custom component f filter data ID generator merge data parse data sequential tasks split text store messages Vector store so we got different Vector store options Astra DB uh Cassandra chroma DB um

You know all these different ones pine cone uh uh cudr uh reddis super base I love super base um they don't have zap which I do like Zep as well um but yeah they do have some of the you know popular ones super Bas quad cudr reddis mongod DB Atlas um chroma DB and then embeddings so hugging face uh mistol aama uh open Ai and then experimental so I was looking at this too this is kind of cool so we got

Agents wow oh this is awesome crew AI agent I'm going to do more videos on that stuff too um crei CSV agent um Json agent SQL agent asra assistant so create assistant. NV uh environment get environment variable listate assistance chains conversational chains loaders utilities so fire crawl API for scraping and whatnot fire Coss scrape API memory oh they do have Zep memory perfect okay so they this is like really a good uh prototype so conditional rer python function uh retrievers Tech

Splitters tool kits and then tools here so a lot of stuff to play around with learn Etc let's just say for example um I go here I click on this and you can I like this too is that you can see um like if I go here code I can literally see this code so from Lang chain open AI embeddings base import open AI embeddings from L like I can see the python code right here I can literally go and run

This and um yeah so it's like gives you the code for python but also the nice gooey interface so that's what I really like about this to um I don't don't quote me on this I don't know I don't don't think flow wise does any some I don't know I like I know that I don't think they do that like in this like flow wise or Diffy do that with uh python I don't think they do that so that's what I

Really like about uh this tool and then and um yeah I'm excited to and then like let's see here so advanced settings default headers default query chunk size so yeah you can play with the settings here there's a lot of different uh customizations options but anyways um I think you guys get the point here oh there's also a playground okay so this is where you would talk to it and then API so this is where you can run C command python

API JS API python code chat widget similar to flow wise and Diffy and just anything in that regard um and then share um you can share the rag or the the workflow of the project with others so other than that guys I I'm very excited to start using this Tool uh I'm probably going to take this and put it into coolify to self-host it um I don't like doing running it the stuff locally too much um just yeah but if you

Do want me to show you guys how to do that in coolify or that's what I've done many videos on coolify if you're not familiar with coolify check out some of those other videos but you can host this in like whatever you really want but uh yeah I'll probably um set this up in cool V and if you want me to make make a video specifically on that showing you just giving you the code for that let me know in the

Comments down below other than that guys like I said if you're new here make sure to smash that like button smash that subscribe button to stay up to date with the daily uploads also too guys if you're new and you haven't already join the stride Community our free community of like-minded AI enthusiasts entrepreneurs developers coders um marketers salespeople Etc check that out Link in the description down below it's 100% free our free Facebook group free Discord Channel we got about 750

800 members in both those combined right now also too guys if you haven't already checked out my book client acquisition code check that out Link in the description Down Below guys other than that um if you have any questions about this video or any ideas or whatever let me know in the comments down below but I will see you in tomorrow's video guys hope you're having a great weekend keep hustling keep grinding and of course guys accelerate your stride take care