Use Claude Code 100% FREE & UNLIMITED with Fast Responses (Omniroute) — Transcript
Full transcript
- 0:00You need to stop paying for Claude code,
- 0:02and instead run it completely for free
- 0:04with 500 of the best AI models on the
- 0:07planet. And in this video, I will show
- 0:09you exactly how you can do that using a
- 0:12tool called OmniRout, which lets you
- 0:14connect Claude code to models like GLM,
- 0:17DeepSeek, Gemini, and even some of
- 0:19Anthropic's top models like Opus 5 and
- 0:22Sonnet completely for free. And at the
- 0:24end, I'm going to put it to the test by
- 0:27building a multi-page website and a
- 0:29full-stack app to see if the results
- 0:31will be the same with Claude's paid
- 0:33tier. So, close all of your other tabs,
- 0:36and let's dive right into it. But before
- 0:37I show you how to set this up, I want to
- 0:39first talk about why would you actually
- 0:41want to stop paying for Claude code?
- 0:43Since Claude code on its own is the most
- 0:46powerful coding agent on the planet. But
- 0:48the negative with Claude code is that
- 0:50your usage tends to run out pretty fast,
- 0:53especially if you're not in the Max
- 0:55plan. You also get the rate limited a
- 0:57lot, and you have to wait hours until
- 0:59your rate limit resets. And even with
- 1:02these two negatives, your monthly bill
- 1:05can still be in the hundreds of dollars.
- 1:07But that's where OmniRout comes in.
- 1:09Because what OmniRout does is it serves
- 1:12as one endpoint that gives you access to
- 1:14over 500 models from 290 different
- 1:18providers. And included in those models,
- 1:21there exist 90 plus free models that you
- 1:24can access through OmniRout 24/7 without
- 1:27needing to put in your card or even sign
- 1:29up for an account. And the best part is
- 1:32that OmniRout, using a clever trick,
- 1:34also allows you to get Opus and Sonnet 5
- 1:37completely for free as well. And don't
- 1:40worry because I will show you this trick
- 1:41in this video. But now you might say,
- 1:43"Well, Damian, that sounds good, but
- 1:45will I not also get the rate limited
- 1:48when I'm using OmniRout?" And the answer
- 1:50to that is a resounding no. And that's
- 1:52because OmniRout was specifically built
- 1:55to fix that problem. Since what Omni
- 1:57route does is it gets a request from,
- 2:00let's say, your cloud code agent, it
- 2:02looks into all the models that you have
- 2:04connected it to, which can be many
- 2:06different models from different
- 2:08companies, and then it picks the best
- 2:10one of these models for the task that
- 2:12you want to do. But, if the top one for
- 2:15the task that you want to do has already
- 2:17hit a rate limit, then it simply moves
- 2:19you down to the second best one, and
- 2:21then it uses that model to complete that
- 2:23task. And the good thing with Omni route
- 2:25is that you can literally connect 12 or
- 2:27even more models at once, which means
- 2:30that once a model hits a rate limit,
- 2:32then Omni route will simply move your
- 2:34request to the next best one. And
- 2:36another feature of Omni route is that it
- 2:39takes your prompt and compresses it, so
- 2:41it spends fewer tokens for the same
- 2:43request. And from tests that I have
- 2:45done, they have found your token saving
- 2:48can be up to 95%, which is honestly
- 2:51massive. And even if you're not
- 2:52interested in using free models, even if
- 2:55you want to use paid models normally, it
- 2:57still makes sense to use Omni route
- 2:58simply because with it you can spend up
- 3:01to 95% fewer tokens. So, now that you
- 3:04understand how Omni route works, let me
- 3:06show you how you can install it and how
- 3:08you can make cloud code use it. Now, to
- 3:10install it, you simply need to go over
- 3:12to omniroute.online,
- 3:14which I will also leave a link for in
- 3:15the description below, and then you need
- 3:17to come over here and click on start
- 3:19free, which is going to bring you over
- 3:21in this tab, which gives you access to
- 3:23all the commands that you need to
- 3:24install it and also run it. Now, the
- 3:26fetch command is this command over here,
- 3:29which is npm install omniroute. So, you
- 3:31will simply need to copy this command
- 3:33and then open up a terminal. It can
- 3:35either be inside an IDE, like Antegraphy
- 3:38or Cursor, or it can even be the native
- 3:41terminal in your machine, but preferably
- 3:44open it up inside an IDE with a coding
- 3:47agent next to it, and you will see why
- 3:48in a bit. Paste in that command and
- 3:50click on enter. And then what this is
- 3:52going to do is start installing Omni out
- 3:55in your machine. And now, all you have
- 3:57to do is to wait 1 minute till it is
- 3:59done. And boom, as you can see, this is
- 4:01just finished installing. Now, if you
- 4:02get some text like this, don't actually
- 4:04worry. This is just a normal
- 4:06installation process. And now, to
- 4:08actually run it, we simply need to come
- 4:09over here and type Omni out and click on
- 4:12enter. Well, what this is going to do is
- 4:14spin up this local host, which as you
- 4:16can see, it also automatically opened up
- 4:19in my browser. But if it doesn't do that
- 4:21for you, all you will simply need to do
- 4:23is to simply come over here and open it
- 4:25up from this local host dashboard URL.
- 4:27And then, once you do that, you will
- 4:29simply need to sign in with your
- 4:30account. Now, you currently don't have
- 4:33an account, so what you need to do is to
- 4:35simply come over here and copy the
- 4:36password that it gives you, which is
- 4:39change me in all caps. So, simply copy
- 4:42it, paste it over here, and click on
- 4:43continue. And then, this is going to
- 4:45successfully log you in into Omni out.
- 4:48Now, make sure you don't copy any spaces
- 4:50in the password, because if you do, then
- 4:52you will not be able to log in. And once
- 4:54you are inside the Omni out dashboard,
- 4:57then it's time to connect all the
- 4:59providers that you want to use. Now, to
- 5:01do this, you simply need to come over
- 5:03here to the connect providers tab and
- 5:05click on providers, which is going to
- 5:07bring you over in this section over
- 5:09here. Well, this might seem intimidating
- 5:11at first, but don't worry, this is
- 5:13actually super simple to navigate.
- 5:15Since, as you can see here, we have all
- 5:17the different providers that we can
- 5:19connect to. We can see some very big
- 5:21names like Kimiko, Antigravity, Amazon,
- 5:24Code X, along with a lot more. Now, to
- 5:27be able to navigate these providers and
- 5:29find the ones that you want, we have
- 5:31this filter section over here to the
- 5:32top, where we can come over here and
- 5:34select for the characteristic of the
- 5:36provider that we want to find. For
- 5:38example, if you want providers that
- 5:41don't have any off, which means that you
- 5:43don't need to create an account to be
- 5:45able to use them, you can simply click
- 5:47on no off and then this is going to
- 5:49return you all those providers. Or if
- 5:52you want providers that have a free tier
- 5:54that gives you free models, you can
- 5:56simply come over here and click on free
- 5:58tier and this is going to tell you all
- 6:00the providers that offer that. Plus, if
- 6:02you want a specific provider, you can
- 6:04simply come over here and search for it
- 6:07through the search bar. For example, if
- 6:08I wanted to find open outer, I will
- 6:11simply come over here and search for
- 6:12open outer and then this would tell me
- 6:15the open outer provider. Where to
- 6:17actually connect to it, I will simply
- 6:18have to click on it and then click on
- 6:21add connection. Now, this is going to
- 6:23ask me for the name. I will leave it on
- 6:25main and then this is going to ask me
- 6:27for my open outer API key. So, to create
- 6:30my open outer key, I will simply go over
- 6:33to open outer, click on get API key and
- 6:35simply come over here and create a brand
- 6:37new key that I'm going to call Damian
- 6:39and then I will simply copy it, come
- 6:41back over to Omni route, paste it in
- 6:43over here and I can also click on check
- 6:45to verify that this is a valid API key
- 6:47and then I have the option, if I want,
- 6:50to only import the free models that open
- 6:53outer gives us access to, which are all
- 6:55these models over here. So, if you want
- 6:57only the free ones, you can simply turn
- 7:00this on and click on save. And what this
- 7:02is going to do is import all the 70 free
- 7:05models that you can access through open
- 7:07outer, which you can actually see and
- 7:09verify that they work from over here.
- 7:11Now, let's say for example, I wanted to
- 7:13see if the Gemma 4 model from Google
- 7:15actually works. All I would have to do
- 7:17is to simply come over here and click on
- 7:18the play button and this would actually
- 7:20send the request over to the model to
- 7:22see if it is actually working. And you
- 7:24will see that if I click on all these
- 7:26models, some of them might actually give
- 7:28me a red X, which means that they are
- 7:30not actually working. For example, this
- 7:32model over here is not working for some
- 7:34reason. Maybe it's because the server is
- 7:36overloaded or maybe it's for another
- 7:38reason, but currently it is not working
- 7:41and I can see it because I got this
- 7:43error message over here. So, if you
- 7:45actually do this with all the models
- 7:46that you imported, you will be able to
- 7:48see which ones are working and which
- 7:50ones aren't, which also lets Omni route
- 7:52know in which models it can actually
- 7:55send your request to. But, since the
- 7:57goal of the video is for me to
- 7:59specifically show you how you can use
- 8:01the best free models inside Omni route,
- 8:04let me actually come over here and give
- 8:06you my own personal suggestions. Now, a
- 8:08very good provider that I would
- 8:09definitely connect, which is Open Code,
- 8:11and specifically the Open Code free
- 8:14provider. Well, as you can see, I can
- 8:16actually click on it, and this
- 8:17automatically enables it since this is a
- 8:20no authentication connector, which means
- 8:22that I don't even need to create an
- 8:23account. As you can see, this gives us
- 8:25access to all of these six models that
- 8:28we can use completely for free, which
- 8:29include a very powerful model like Deep
- 8:32Seek V4 Flash. Now, another very good
- 8:35provider that I would definitely advise
- 8:37you to connect to is NVIDIA. Since
- 8:40NVIDIA actually gives us access to more
- 8:43than 30 free models, and to connect to
- 8:45it, we simply have to come over here,
- 8:47click on add, and then we need to give
- 8:49it our NVIDIA API key, which we can get
- 8:52by going over to the NVIDIA platform.
- 8:54The link for it will also be in the
- 8:55description below, then simply signing
- 8:58up for an account, and then moving over
- 9:00in the API key section, where we will
- 9:02then be able to create an API key,
- 9:04exactly as I'm doing right now, that I
- 9:06will then be able to copy, come back
- 9:08over to Omni route, paste the thing, and
- 9:10I will also select to only import the
- 9:12free models, and I'm going to click on
- 9:14save. And boom, this imported all the
- 9:16free models that NVIDIA gives us access
- 9:18to. As you can see, these are literally
- 9:20all the models over here, and this
- 9:22includes some very powerful models like
- 9:24Mini Max M3, Kimiko 2.6, and even GLM
- 9:295.2. So, now let me also close this
- 9:32because now I want to show you two
- 9:34providers that actually give you access
- 9:36to Anthropic paid models like Opus 5 and
- 9:40Opus 4.8 completely for free. And the
- 9:43first provider is called Agent Router,
- 9:46which is this provider over here. Which
- 9:49if I actually click on, then I'm able to
- 9:51come over here and create a connection,
- 9:53which you can access by going over to
- 9:55agentrouter.org,
- 9:57creating an account with GitHub, and
- 9:59then clicking on get key. And the good
- 10:01thing with Agent Router is that they
- 10:03give you $100 and $75 worth of free
- 10:07credits that you can spend on Opus 5 and
- 10:10Opus 4.8. Now, I have made a dedicated
- 10:12video about how to create an Agent
- 10:14Router account. You can go and watch it.
- 10:16I will put it somewhere over here. But
- 10:18for now, all we have to do is to simply
- 10:20create an API token, which I will do by
- 10:22clicking on create. I'm going to give
- 10:24this a name, and then I will simply
- 10:26click on create, which is going to
- 10:28generate my API key that I will then be
- 10:30able to copy, come back over to Agent
- 10:33Router, I'm going to paste it in, and
- 10:34I'm going to click on save. And this is
- 10:36going to import all the models from
- 10:38Agent Router, where as you can see, I
- 10:40have models like Opus 4.8, Opus 5, and
- 10:44GPT 5.6 all. And if I actually test them
- 10:46out, I'm going to see that they are
- 10:48working correctly. And that now means
- 10:50that we can access even the most
- 10:51powerful Anthropic models like Opus 5
- 10:54and Opus 4.8 completely for free through
- 10:57Agent Router. Since again, Agent Router
- 10:59gives us $175 worth of credits when we
- 11:03first sign up. Now, another connector
- 11:05that gives us Anthropic models
- 11:07completely for free is Kiro AI, where
- 11:09you can simply search for it, and then
- 11:11click on add connection. Then you simply
- 11:13need to click on I understand and
- 11:15continue, and then you can select how
- 11:17you want to connect to your Kiro
- 11:18account. Now, I'll simply select
- 11:20continue with Google, and then I'm going
- 11:22to click on approve, which is going to
- 11:23authorize my Kiro account and
- 11:25successfully connect to Omni Router.
- 11:27Which will then mean that we will be
- 11:28able to use all all different Anthropic
- 11:31models like Sonnet 5 and Opus 5 from
- 11:34within our Kira account. Now, I know
- 11:36that this process might seem a bit
- 11:37boring and repetitive right now, but the
- 11:40good thing with this is that you only
- 11:42have to set it up once and then OmniOut
- 11:44will basically automatically keep
- 11:46switching between all these providers
- 11:48when you hit a rate limit in one of
- 11:51them, which means that you will
- 11:52constantly be able to code with all
- 11:54these powerful models without having to
- 11:56stop and wait for your rate limits to
- 11:58get reset. So, what I actually strongly
- 12:00advise you to take your time and connect
- 12:02to all the providers that you want. Now,
- 12:05I'll show you the ones that I personally
- 12:07think are the best when it comes to
- 12:08getting the best models completely for
- 12:10free, but if you want you can connect to
- 12:12other providers like Under Gravity
- 12:14through the CLI, Amazon, and a lot more.
- 12:17But, I do want to make one disclaimer
- 12:19and that is when you're connecting
- 12:21through providers like Under Gravity via
- 12:24authentication in your account and not
- 12:26through an API, then some of these
- 12:28companies don't actually allow that,
- 12:30which does run a small risk of your
- 12:33account being banned and I do want to
- 12:34point that out because some people's
- 12:36accounts have been banned before, but
- 12:38that was when they were having very
- 12:40heavy usage. And again, the ban risk
- 12:42only exists on the ones that you have to
- 12:44connect through authentication. So,
- 12:46anything else where you don't need
- 12:48authentication or you connect through an
- 12:50API, you're 100% safe and you can 100%
- 12:53safely connect it to your OmniOut
- 12:55account. And now that we have connected
- 12:57all of our providers, it's time where I
- 12:59show you how you can connect OmniOut to
- 13:02Claude Code so you can actually use all
- 13:04the free models that they offer. Now, to
- 13:06make this happen, we first have to
- 13:08install Claude Code in our machine. Now,
- 13:10to do this, we need to go over to the
- 13:12Claude Code docs and use the command for
- 13:15our operating system. So, if you are on
- 13:17macOS or Linux, use this command over
- 13:19here and if you're on Windows, then use
- 13:21the command at the bottom. Now, since
- 13:23I'm on Mac, I will simply copy this
- 13:25command and then you need to open up a
- 13:27brand new terminal tab, exactly as I did
- 13:29right now, and simply paste in that
- 13:31command and click on enter. Because what
- 13:33this is going to do is start installing
- 13:35Cloud Code in your machine, which as you
- 13:38can see it just finished doing. And now
- 13:40to launch it, all we have to do is to
- 13:42type in Cloud and then this is going to
- 13:44launch Cloud Code in our machine. But
- 13:46currently I'm using Opus 5 from my
- 13:48Anthropic subscription, which we need to
- 13:51switch and make Cloud Code use the
- 13:53models that we just connected in
- 13:55Omniroute. Now to do that, you will need
- 13:57to use this prompt that I have left a
- 13:59link for in the description below. And
- 14:01what this prompt basically does is it
- 14:03tells your coding agent the exact
- 14:06changes that it needs to make inside the
- 14:08Cloud settings folder to actually enable
- 14:11to be able to use Omniroute. Now you
- 14:13don't actually need to understand how
- 14:14this actually works and all you have to
- 14:16do is to simply come over here and copy
- 14:19the full prompt that has all the
- 14:21instructions that your coding agent
- 14:23needs and then you can either paste this
- 14:26inside the coding agent assistant of
- 14:28your ID or if you want you can even
- 14:30paste it inside the Cloud Code instance
- 14:33that you just created. Since I also want
- 14:35to show you how you can set this up
- 14:36completely for free as well, I'm going
- 14:38to do this with a 100% free Antigravity
- 14:41agent, which I will also leave a link
- 14:43for in the description below. So I'll
- 14:45simply send over this prompt so my
- 14:47Antigravity agent can go change my Cloud
- 14:50settings and make it so Cloud Code
- 14:52doesn't route the request to my Cloud
- 14:54Max subscription and it actually routes
- 14:56it to Omniroute. The only caveat to this
- 14:59is that you need to have a live terminal
- 15:01with Omniroute up and running. So if you
- 15:04close your previous terminal, make sure
- 15:06you spin up a new one because Omniroute
- 15:08needs to be live in your local host in
- 15:11order for this to be able to work. But
- 15:12since I already have it open, I don't
- 15:14need this brand new one. And now all I'm
- 15:16going to do is I will simply wait till
- 15:18my Antigravity agent is done, it makes
- 15:21the changes that it needs and then I'm
- 15:22able to access all the Omni route models
- 15:25that are connected from inside Cloud
- 15:27Code. But, while your coding agent is
- 15:29setting this up, it is going to ask you
- 15:31what model you want to use, and this
- 15:33will return you a bunch of options, like
- 15:36the auto best for coding, which
- 15:37specifically picks the models that are
- 15:39best for coding, the premium coding
- 15:41routing, the auto cloud/Opus, which
- 15:44routes only through Cloud Opus models,
- 15:46along with a bunch of more options. Now,
- 15:49these are basically the pre-configured
- 15:51templates that Omni route has created,
- 15:52which see what models you have
- 15:54connected, and it picks the best of
- 15:56those for the specific goal that you
- 15:57select. Now, since we are setting this
- 16:00up with completely free models, I will
- 16:02select the auto best free, which will
- 16:05allow Cloud Code to use the best free
- 16:06models that we have connected to Omni
- 16:08route. But, if you want models from just
- 16:11a specific provider, you can simply come
- 16:13over here and say, "Hey, I only want
- 16:15models from Open Router, let's say." And
- 16:18then, our coding assistant will make it
- 16:21so Cloud Code can only access the models
- 16:23that you have connected to Omni route
- 16:25from only Open Router. But, since most
- 16:27of you want the best free models from
- 16:30all the providers that we selected,
- 16:31simply select auto best free, and then
- 16:33click on submit, and then simply let our
- 16:36coding agent run and set this up for us.
- 16:38And boom, as you can see, that's exactly
- 16:40what our Under Gravity agent did, and it
- 16:43is telling us that it also verified the
- 16:44connection. Plus, here it is also
- 16:46showing us the block of code that it
- 16:48changed. Now, for the most technical
- 16:50people, you will be able to see that
- 16:51this is basically changed the provider
- 16:53to a local host, and then it also
- 16:55selected the specific mode that we want,
- 16:57which is auto best/free. But, you don't
- 17:00actually need to understand how this
- 17:01works, since the AI understands it for
- 17:04you. And now, if I actually open up a
- 17:06brand new terminal and I type in Claude,
- 17:09this is going to spin up another Claude
- 17:11Code instance. But, as you can see, this
- 17:13time it is using the auto best free mode
- 17:17from Omni route. And we can verify this
- 17:20by typing the {slash} stats command. We
- 17:22will be able to see that this is
- 17:24pointing to a local host that is live
- 17:26right now, which is basically Omni
- 17:27route. And over here we can see the
- 17:29model, which is this one. And just so
- 17:30you guys can see that this actually
- 17:32works, if I come over here and say
- 17:34hello, we are actually going to get the
- 17:36response back from it that says "Hello,
- 17:38I'm Kito." Since it currently seems to
- 17:40be using the free Kito models that we
- 17:42connected, and it is asking us what we
- 17:45want to
- 17:46on. Now, let's say for example that we
- 17:47wanted to build a full multi-page
- 17:49website. All we would have to do is
- 17:51simply come over here and say, "Hey,
- 17:52what I want you to do is to go over and
- 17:54build me a multi-page functioning
- 17:56website for a dental clinic. Make it
- 17:58look modern and nice." As you can see,
- 18:00pretty vague prompt, but I specifically
- 18:02said it that way to see what Claude
- 18:05Code, using the best free models, will
- 18:07be able to create. And I'll simply send
- 18:09over this prompt so Claude Code can
- 18:10start building out this multi-page
- 18:12website for me. But while he's doing
- 18:14that, I want to talk to you about
- 18:16probably the most important function of
- 18:18Omni route, which is the custom combos
- 18:20that you can create. Now, to access the
- 18:22combos, simply come over here to the
- 18:25Omni proxy section, and then you will be
- 18:27able to see the combos page over here.
- 18:30And if you actually click on it, then
- 18:32this is going to show you this interface
- 18:33over here. Where the first thing that
- 18:35you can see is the 17 pre-built
- 18:38templates that Omni route has already
- 18:40created for us. And if I actually click
- 18:42on this, which is going to open it up,
- 18:44we can see what these templates actually
- 18:46are. Here we have the Auto Best Coding,
- 18:48which is a template that selects the
- 18:50best coding models that we have
- 18:52available. Then we have the Auto Best
- 18:54Vision, which selects the best ones for
- 18:55vision. And we also have the Auto Best
- 18:58Free Combo, which is the combo that
- 19:00we're actually currently using that
- 19:02selects the best free model for the task
- 19:05that we told Claude Code to complete.
- 19:07But the good thing with combos is that
- 19:09you can create your own. And you can do
- 19:11this very easily by simply coming over
- 19:13here and clicking on create your first
- 19:15combo. Now, the first thing that you
- 19:16will have to do is to simply give it a
- 19:18name. I'm going to call this one Damian,
- 19:20for example, and then we can select
- 19:22whether we want to use a quick template.
- 19:24For example, we can either select free
- 19:26stack, which would only be using free
- 19:28models, or we maybe want to select high
- 19:31availability, so we only send models
- 19:33that have high availability, or models
- 19:36that are pretty cheap, or balanced load,
- 19:38or any of these other templates. And
- 19:40now, I will personally choose to go with
- 19:42a free stack for this example, and then
- 19:44we'll click on next, which is going to
- 19:45bring me over in the next section, where
- 19:47what I need to do is to simply select
- 19:49the provider that I want, and then I can
- 19:52also select the specific model from that
- 19:54provider. For example, if I want the
- 19:56first model to be from Invidia, and then
- 19:58let's say I want to specifically GLM
- 20:005.2, I can simply select it, and this is
- 20:03going to be added in my combo. And then,
- 20:05if I click on add step, then this is
- 20:07going to add it over here in my combo
- 20:09models, where as you can see, currently
- 20:11is in the eighth place, but if I want to
- 20:13move it up, I can simply click on the
- 20:15move up button, and this is going to
- 20:17move the model that I just added, which
- 20:19is GLM 5.2 from Invidia, and make it the
- 20:22first option for that combo. And then,
- 20:25if that model hits a rate limit or is
- 20:27very late to respond, then my AI agent
- 20:30is going to call the Gemini 3.5 flash
- 20:33model. But if I don't actually want this
- 20:35to happen, and let's say I wanted the
- 20:36Kiro model after that, I could simply
- 20:38select it. Or maybe let's say I wanted
- 20:40the Kiro Sonnet 4.5 to be the first, I
- 20:43can simply move it over to the top, and
- 20:45I have GLM 5.2 underneath it. So, we can
- 20:47set the exact order of models that we
- 20:49want while recording or performing any
- 20:52other task. Plus, we don't need to just
- 20:54have eight models over here. We can add
- 20:56in as many as we want. And all we have
- 20:58to do is to simply select the provider,
- 20:59then also select the model. Let's say,
- 21:01for example, I wanted Claude Sonnet 5,
- 21:04and then I will simply click on add
- 21:05step, and this would add this model on
- 21:08my combo where if I wanted then I could
- 21:10move it at the top of the combo. And
- 21:12once I'm done selecting the models, then
- 21:14I can simply click on next, which is
- 21:16going to bring me over in this tab over
- 21:17here where I can select my routing
- 21:19strategy. I can either do priority,
- 21:21weighted, or a round robin. Now, if you
- 21:24want to hold the priority that you set
- 21:26in the previous section, you will need
- 21:28to select priority in this tab. And then
- 21:30you can also fill in some more advanced
- 21:32settings if you actually want. I will
- 21:33choose to skip them for now. I will
- 21:35click on next, and then I can review
- 21:37everything that I have selected and
- 21:39click on create combo. And boom, our
- 21:40combo has been successfully created. And
- 21:43now if I scroll down here, I will be
- 21:45able to see it with the name Damian and
- 21:47the exact settings that I selected. But
- 21:49before we do that, I want to talk a
- 21:51little bit about the sponsor of today's
- 21:53video, which is me and my own AI file
- 21:56organization app that is called File
- 21:58Act. Because what File Act does is it
- 22:00allows you to automate the file
- 22:02organization in your PC by simply giving
- 22:05one prompt to AI, and then it doesn't
- 22:07organizes the files and folders that you
- 22:09want for you. So you never again have to
- 22:11spend 10 or 20 minutes trying to find
- 22:14files and folders in a very messy
- 22:16desktop. So if you would be interested
- 22:17to find out how it works, along with
- 22:19claim a 10-day free trial, so you can
- 22:22try it out completely for free, then
- 22:24make sure you click the second link in
- 22:25my description, you download the app,
- 22:27and you start your 10-day free trial.
- 22:29And now if I wanted to make Cloud Code
- 22:31use my custom combo, all I would have to
- 22:33do is to simply copy this prompt, and
- 22:35then either come back over in my Under
- 22:38Gravity Agent, paste in the name of the
- 22:40combo and say, "Hey, this is my brand
- 22:41new combo that I just created. Make
- 22:43Cloud Code use that one instead of the
- 22:45current one." and send over the prompt,
- 22:47or I could even open up another Cloud
- 22:50Code instance, say the exact same thing,
- 22:52click on enter, and I would have Cloud
- 22:54Code go inside the settings and change
- 22:57it by itself, which it just finished
- 22:58doing. The only thing that I had to
- 23:00change is again paste the prompt that I
- 23:02showed you at the start. So, Claude Code
- 23:04also has context for how we can do that.
- 23:07But, once I did it, it successfully went
- 23:09over it and changed that for me. And
- 23:10now, if I open up a brand new terminal
- 23:13and I type in Claude, you guys will be
- 23:15able to see that now this is using the
- 23:17Damian combo that I just created. And if
- 23:19I send "Hello, what model am I using?",
- 23:22you will see that it is going to tell me
- 23:24that I'm using Damian, which is the
- 23:25combo that we just created that as a
- 23:27first option is using Claude Sonnet 5.
- 23:30And while this Claude instance was
- 23:32making that change, my other Claude
- 23:34instance that was building the dental
- 23:36clinic website finished executing and it
- 23:39spun it up in a localhost URL, which I
- 23:41can now open and actually take a look
- 23:43at. So, as you can see, this is the
- 23:44website over here. Here we have the hero
- 23:46section. Then we have some information
- 23:49about the clinic. Then we can see all of
- 23:51their services with some nice
- 23:52animations. And then we have a simple
- 23:54CTA. Plus, we also have all the other
- 23:56pages like services, which is a
- 23:58completely unique page as you can see
- 24:00over here. Then an about us page as
- 24:02well, which is completely unique and it
- 24:04has some information about the company.
- 24:06Then we have an our team page, which
- 24:08introduces us to the members of that
- 24:10dental clinic's team. And a contact us
- 24:12page, where we can fill in our
- 24:14information and then dental clinic is
- 24:16going to contact us, which is exactly
- 24:18what we asked Claude to do. So, you guys
- 24:20can see how powerful the Claude Code
- 24:22plus Omni route combo can be. Since we
- 24:24can literally have access to all these
- 24:25different powerful models from all these
- 24:28different providers completely for free.
- 24:30Now, if you want more dedicated help
- 24:32with setting this up, then make sure you
- 24:33click the top link in my description and
- 24:35you come and check out my school
- 24:36community, where I help people just like
- 24:38you learn and solve any issue that they
- 24:40have when it comes to AI. Plus, I also
- 24:42help them build their own AI SaaS and
- 24:45scale it to over $1,000 in MRR, just
- 24:48like as I did with my own app called
- 24:51File Elect. So, if you would be
- 24:52interested in that, along with getting
- 24:54dedicated one-to-one access to me, then
- 24:56make sure you click the top link in my
- 24:58description and you comment cycle my
- 24:59school community. And if you want to
- 25:01automate file organization in your
- 25:03machine, then make sure you click the
- 25:04second link in my description and you
- 25:06also check out File Elect. But also you
- 25:08do things that you're going to enjoy
- 25:10this video, in which to show you another
- 25:11way with which you can use cloud code
- 25:13completely for free. So make sure you go
- 25:15and watch that next and I will see you
- 25:17over there.
About this transcript
This page contains the full transcript of Use Claude Code 100% FREE & UNLIMITED with Fast Responses (Omniroute) by Damian Malliaros, generated from the public captions YouTube serves with the video. The transcript has 5,329 words across 713 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.