Get 7.4 Billion Free Tokens for Claude Code (Full Setup) — Transcript
Full transcript
- 0:00So guys, today we're going to see how
- 0:02to get 7.4 billion AI tokens per month
- 0:05for free. In this, we will connect a
- 0:08total of 34 free providers. 635 free
- 0:11model endpoints and 474 model families.
- 0:16This will also include chat, embedding,
- 0:18and video models. So in this video, we
- 0:20will look at how we can configure this.
- 0:22We will use smart fallback routing for
- 0:25all models, meaning we will switch
- 0:27models automatically. We will handle
- 0:30rate limiting automatically, and most
- 0:32importantly, we can connect it to any
- 0:34coding agent like Claude Code. So in
- 0:37this video, we will integrate it with
- 0:38Claude Code and see how it works. So
- 0:40without wasting any time, let's start
- 0:42the video. So for that, you need to go
- 0:44to your favorite browser and search for
- 0:47free-llm-api.com, okay. So in this, we
- 0:50can see that the providers have
- 0:52increased; it now has 969 models and 62
- 0:54providers, which have updated to be
- 0:57much higher. They were fewer before,
- 1:00and we will get 7.4 billion free tokens
- 1:02per month in this. And frontier models
- 1:05that we will get for free. Out of which
- 1:08, these frontier models are right in
- 1:09front of you. In this, Gemini 1.5 Flash
- 1:13, GLM 4, Qwen 2.5, DeepSeek V3, and
- 1:17Minimax are all models you can see here
- 1:21. Quite excellent frontier models are
- 1:23available for you to use for free in
- 1:25this. Of course, there is a limit, but
- 1:28the smart fallback routing will ensure
- 1:31we don't feel that limit at all. From
- 1:34here, we need to click on the GitHub
- 1:36option. As soon as you click on the Git
- 1:38option, a Git repository will open in
- 1:40front of you. And here you can see that
- 1:43it already has 36,000 stars. And this
- 1:47is an open-source project that provides
- 1:49us with a service where we can connect
- 1:52all endpoints together and use free
- 1:55LLMs. Here we can see the supported
- 1:57providers. Like Google, Groq,
- 2:00OpenRouter—many excellent providers
- 2:02are available to see in this. So, to
- 2:05set this up, we need to click on Quick
- 2:07Start here. This is the URL you get.
- 2:11For this, you will require Docker. If
- 2:13Docker is installed on your system, you
- 2:15can install it directly with this
- 2:17command. But I will tell you the
- 2:19simplest way to use it. If you are on
- 2:22Windows, you can click here to download
- 2:25the exe file. But I am currently using
- 2:27a Mac. So I will click here for Mac OS.
- 2:32And as soon as you scroll, after
- 2:34scrolling, you will see the DMG file
- 2:37here. If you are also using a Mac, you
- 2:39can download this DMG file. Otherwise,
- 2:42you can also see EXE files here. If you
- 2:45are a Windows user, download the EXE
- 2:47file. It's a very simple process. You
- 2:49can install it easily. So if you are
- 2:51using an M-series, meaning an M-series
- 2:53MacBook. Then you can use the ARM 64
- 2:56version. So I am currently using an M5.
- 2:59So for the M5, I will download the ARM
- 3:0164. So I'll wait for it to finish
- 3:03downloading. After downloading, you
- 3:07need to open it and drag it into the
- 3:09Applications folder. And as soon as you
- 3:12hear this sound, your software will be
- 3:14successfully installed on your system.
- 3:16Now I will open the Free LLM API here.
- 3:25There are three steps to get started
- 3:25with this. Add provider keys, send a
- 3:28test request, and connect your app. So
- 3:30in this, what we need to do is add API
- 3:32keys from multiple providers. We get a
- 3:35playground option in this. From where
- 3:37we will test the request to see if it
- 3:39is working or not. Then we can directly
- 3:41connect any app, like any of our coding
- 3:44agents, with it. In today's video, I
- 3:46will show you how to connect and use it
- 3:48with Cloud Code. So here you can see
- 3:51that in models, we will get chat,
- 3:53embedding, image, video, audio, search,
- 3:55and fused models. And here we get the
- 3:57playground option, which is currently
- 3:59empty. We will add to it. So here we
- 4:02can chat using the keys that we are
- 4:04going to create. In this, we get to see
- 4:07agents that we might want to use it
- 4:10with. Like, we will use it with Cloud
- 4:12Code. In this, we can connect and use
- 4:15it with Codeium, Cline, Continue,
- 4:17OpenCode, and multiple other agents. We
- 4:20can also use it with Cursor. In
- 4:22analytics, we get an entire tab for
- 4:25this. Where we will get analytics for
- 4:27usage, requests, input-output tokens,
- 4:30how much we are using, and how much is
- 4:32being spent, all of that we will see
- 4:34here. Alright? And in the premium
- 4:36section, we will get the options to buy
- 4:38their plans and such. All right? So now
- 4:40let's go to models and I'll click on
- 4:42add key. I'm clicking on add key.
- 4:45Select provider, here you can select
- 4:47any provider you want. I will select
- 4:49Google AI Studio first here. And here I
- 4:52search for Google AI Studio. I will
- 4:56open Google AI Studio here. From here,
- 4:59I click on API key. Get API key,
- 5:02continue, create API key. From here, I
- 5:07click on create key. The key has been
- 5:08created successfully. I copy the key
- 5:10from here and go back into it. I paste
- 5:13the key here. The label is optional.
- 5:17You can provide it here if you want.
- 5:18Otherwise, you don't have to. I will
- 5:20write Google here. I click on add key.
- 5:24Here you can see that 15 models are
- 5:27selected for us. We will get this many
- 5:30models inside this. Like 3.5 Flash, 3.6
- 5:33, 3.7, 3, and even 3.8 will be
- 5:35available in this. So after this, we
- 5:38have to click on save. After clicking
- 5:41on save here, we have to click on check
- 5:43now here. After clicking on check now,
- 5:45it has become healthy here. Earlier,
- 5:47there was an issue here. All right? We
- 5:48have successfully added the first one.
- 5:51Now, I can go to the playground, sorry,
- 5:54to the models, and show you that we
- 5:56have received 96 million tokens for now
- 5:59. 96 million tokens, and we can use
- 6:04these. Now I click on open playground
- 6:07and give it a prompt by clicking here.
- 6:10It is selected to auto. This will
- 6:12automatically generate a response for
- 6:14us on auto. It can choose any model. It
- 6:16has used 3.5 Flash. And it has
- 6:19generated a response for us in this
- 6:22many milliseconds. Now I go back into
- 6:25models. I go into keys. And I click on
- 6:28add key again here to select the
- 6:29provider. There are multiple providers
- 6:32here. You must be able to see them here
- 6:33. Select any provider you want and keep
- 6:35adding them into it. All right? I am
- 6:37adding Open Router to it now. So now I
- 6:40will create an API key for Open Router.
- 6:44I come to the Open Router website. And
- 6:46obviously, I will have to sign up here.
- 6:49Okay. So, I have successfully created
- 6:51the account. And I click on API keys.
- 6:54Okay. So, here I click on the new
- 6:55button. I define a name for it. And the
- 6:58expiration will be one hour
- 7:01automatically. And No limit, and I
- 7:06click on create. Our key has been
- 7:09created here. I copy it from here. And
- 7:12I paste this key here. And here I write
- 7:19OpenRouter. And you can see guys, we
- 7:22have received so many models in this.
- 7:25So, I will save it from here. After
- 7:27saving, I will check it again from here
- 7:29. And its status has become healthy
- 7:32again. In this same way, you can select
- 7:34multiple models inside it. And here,
- 7:36you have obtained 102 million tokens.
- 7:39So, in this way, you connect multiple
- 7:42models. Like, you can connect your
- 7:43providers from here. Add all the models
- 7:46from all the multiple providers you see
- 7:48here. Like, there is Grok. There are
- 7:50quite a few in this. You should try
- 7:52them all once. Add as many as you can
- 7:54inside this. I am not connecting all
- 7:56the providers. But to show you how to
- 7:59use it, I have connected two providers
- 8:01to try it out. And we need to use this
- 8:04in Continue Code. Click on Agents. Here
- 8:08we will find the Continue Code option.
- 8:10We are seeing an automatic setup inside
- 8:13this. This one. Okay? So, I copy this.
- 8:16Uh. I will paste this command inside
- 8:18the terminal. And I press enter. I
- 8:20press Y here. I press enter. So, it is
- 8:23asking for the unified API key. So, we
- 8:26need to click on show API key and copy
- 8:30the unified API key. And I paste it
- 8:34here and hit enter. Continue settings
- 8:38dot json, I press enter. And you can
- 8:41see here that my API key, base URL, and
- 8:44everything has been selected
- 8:45automatically. So, before this, you
- 8:48must have Continue Code installed on
- 8:51your system. Only then will it work.
- 8:53Okay? So now, I will open Anti-Gravity.
- 8:57I will close this for a moment. And so,
- 8:59I open the terminal here. And I will
- 9:02write Claude. After writing Claude, yes
- 9:05I trust this folder. And here you can
- 9:07see that the API is on auto-use billing
- 9:10. I will click on slash status here. So
- 9:14here you can see all the information is
- 9:17visible to you. Okay? So by clicking on
- 9:20its button, I will send a 'Hi' message
- 9:22to it here. So let's see what kind of
- 9:24response we get. Since you have given
- 9:26the prompt for the first time, it might
- 9:28take a little longer. And also, after
- 9:31clicking on Playground, as soon as you
- 9:33give it a prompt, I'll type 'Hi, how
- 9:35are you' again. As soon as you provide
- 9:39the prompt, you will be able to see all
- 9:42its requests in the Analytics section.
- 9:45I have made 11 requests so far. And
- 9:47these are the input tokens. And look,
- 9:50we can check the usage and everything
- 9:53here. And in the Playground, you get to
- 9:56see many other settings as well. So
- 9:59okay. So here we got the response after
- 10:02a minute because it took quite some
- 10:04time. If you use it for the first time,
- 10:06it might take some time. So let's ask
- 10:10this question, "Which model are you
- 10:12using?" It says auto-select; the model
- 10:16here is on auto-select, and as you can
- 10:19see, I would recommend you to use auto.
- 10:24You have Haiku, Auto, and if you need
- 10:27to do high-level thinking work, you can
- 10:31use Opus. You can use Sonnet when you
- 10:34need to do normal tasks. And if you
- 10:36have very easy, simple tasks to be done
- 10:39, you can use Haiku. So basically, you
- 10:42can use it this way. It was a very
- 10:45simple process. And obviously, you will
- 10:48get limits within this. And the more
- 10:51providers you use, the more you can add
- 10:53multiple providers to this, which will
- 10:55give you access to multiple models. So
- 10:57as soon as we come to the website, we
- 11:00come to keys here. So here we get to
- 11:04see "Add Key" for providers. So there
- 11:07are many providers in this. So you can
- 11:10use any of them. Add it. In the
- 11:12Playground, you have it selected on
- 11:14auto. Here, you can also use your
- 11:15models according to your preferences.
- 11:17Whichever model you want to choose, you
- 11:19can select it and use it. You will also
- 11:22find many other settings and things
- 11:24here. So, this is what you will get to
- 11:27see inside. And if you are new to the
- 11:29channel, make sure to subscribe. And do
- 11:31check out the channel once. I have made
- 11:33a lot of videos for you on YouTube.
- 11:35Some really great ones. So do check
- 11:36them out. And see you in the next video
- 11:38. Until then, bye bye.
About this transcript
This page contains the full transcript of Get 7.4 Billion Free Tokens for Claude Code (Full Setup) by Verma Tech icon, generated from the public captions YouTube serves with the video. The transcript has 1,971 words across 267 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.