Quasar Alpha + Cline: New Stealth Model Beats Sonnet 3.7, R1, & Gemini 2.5! (Fully Free)

On the main channel today, I dropped a video covering two mysterious new AI models. One being called the Night Whisper and the other called Quazar Alpha. I highly recommend that you take a look at these videos as well as these models because they're insanely good at coding and could very well be the next hidden state of the art model of next month. Now, I won't dive too much into Night Whisperer today cuz unfortunately it has been taken down from Ella Marina, but we're going to be shifting our focus to Quaazar Alpha, which is still available through Open Router. Now, these models were stealth dropped through Open Router, which is an AI model router, and there was no official announcement, which is kind of wild cuz you have this mysterious 1 million token context middle model that just randomly appears. And this is something that not only beats Cloud 3. 7 Sonnet as well as Gemini 2. 5 Pro on multiple benchmarks, but it's also running four times faster than these other models. And the craziest part is nobody knows which lab actually created it. And as of now, it's completely free to use off of a open router. My personal guess is that it might be a Google model that they have been quietly testing cuz this is something that they have a knack of doing where they test these different experimental models. They're also known for randomly naming these test models and we all know that this model has a 1 million context window. So, it kind of fits their theory cuz all of their models have that 1 million context window. But obviously, this is speculation. I don't know what the true model is behind this. But overall, it is something that can beat Claw 3. 7 Sonet as well as Gemini 2. 5 Pro on multiple key metrics. It delivers the same performance as Quen 32B as well as DeepSeek R1. It's not as big as the other proprietary models. It's actually slightly smaller, which is kind of surprising. But in coding, it absolutely shines where it nearly matches 03 Mini and Deepseek R1 in Python as well as Bashtask. The speed difference is immediately noticeable, and this is where I would like you to take a look at this video. Now, I'm not lying to you, but take a look at Quaazar alpha in cursor. This is something that isn't sped up, so just keep that in mind, and you will see how rapid it is in terms of generating code. You can see that it generated almost everything from the prompt that was sent in. And it could do it rapidly, which is just insane. Now, this right here is kind of interesting cuz the actual format ID of the model represents something similar to Open AI if you are to compare the two model IDs. But to Google, it's definitely something that's really different as well as with Mistral. So that's why some people might be also saying that this could be an OpenAI model. But regardless, you can definitely go ahead and access it today completely for free through open router. You can head over to Open Router. Make sure you create an account. You can either go ahead and chat via open router or you can go ahead and access it via API through client. Now, what we're going to do next is showcase how you can set this up within client and then showcase a couple of tests that will get us a better understanding of what you can do with this model. But for the people who do not know what client is, it's a autonomous coding agent directly within your IDE that's capable of creating, editing files, executing commands, using your browser, and so much more autonomously. To get started, you'll need your ID. In this case, I'm going to be using Visual Studio Code, but you can use cursor or Windsor or any other ID. Then you need to make sure that you have client installed within that ID. So, I'll leave a link to this as well as all the other links that I use in today's video in the description below. And what you can do is simply go ahead and install this completely for free. This is the client extension that you can access within your IDE within the lefth hand panel. Now what I want you guys to do is head over to the settings tab and then click on API provider and then go ahead and find open router. Now I want you to go back into open router and I want you to generate a new API key which is completely free and you can create an account completely for free as well. Once you have gotten the key, paste it over here. You can do so by clicking on create API key and then I want you to go ahead and search up quazar uh alpha and then you can go ahead and select this start generating anything. So let's get straight to it. This is where I'm providing this screenshot of my Twitter page. I'm going to have it clone it. So let's go over and send uh this prompt to clone this entire website. It's going to go ahead and rapidly work on building it out. And you can see how fast this model is. it goes right away and starts working on generating the HTML file. And obviously once it has finished it, I'll be then opening it up to showcase what it was capable of generating. And just take a look at that. It has fully generated the front end of this Twitter page that I have.

And it actually looks really accurate. Obviously, it's not spaced correctly in the middle, but it did get the job done. It did focus on extracting all the components based off the screenshot I gave, and it accurately generated the icons for each of the different tabs. So, this is quite impressive to see, actually. Next up, I'm going to do something quite vague. I'm going to go ahead and create an interactive simulation of space. So, let's go ahead and send in this prompt, and let's see what it's capable of generating. Now, what I really love about this model is how fast it is. And just like that, it generated this beautiful simulation of space. And you can see all these different stars twinkling and they're slightly moving in a different direction. This is just amazing to see how fast this model is in terms of coding and how it can rapidly work on any task that you give it. Next up, I'm requesting it to create me a SAS landing page. Now you can see it is going to take my prompt and going to start working on generating it quite quickly. And this is why I think many people would want to use this model. Obviously it being free at the moment is a great uh additive. And there we go. This is what I had outputed with this quazar model. But I will say one thing, it doesn't look too great and it doesn't do the same sort of justice as the Gemini 2. 5 Pro or the Claw 3. 7 Sonnet on this prompt to create a SAS landing page. But overall, the other prompts that I had sent in definitely did a greater job at it. But that's basically a gist for today's video on this new Quazar Alpha model. It's truly impressive. It's rapid and it has the potential to be one of the best models that's available in regards to coding. I'll leave all these links in the description below so that you can easily get started. But huge props to whoever has developed this model. I personally think it's a Gemini model that might be slightly smaller and it might be not at the same level as all the coding models that are out there, but it is a lightweight model that I think could be something that Google might have developed. But that's just my two cents. Let me know what you guys think. I'll leave these links in the description below. Subscribe to the newsletter. Follow me on the Patreon. Make sure you subscribe to or not subscribe but follow me on Twitter. And lastly, make sure you guys subscribe to the YouTube channel. Make sure you take a look at our previous videos because there's a lot of content that you'll truly benefit from. But with that thought, guys, have an amazing day. Spread positivity and I'll see you guys fairly shortly. He suffers.