Optimus Alpha: NEW Stealth LLM o4 Mini?! Beats Claude 3.7 Sonnet and Gemini 2.5 Pro (FULLY FREE)

I don't know about you, but it feels like we're officially entering a new era of stealth AI model drops. Cloaked releases that appear out of nowhere, oftenly anonymous. These quiet launches seem to be designed to generate hype for models we have already suspected are on the horizon, like Gemini 2. 5 Flash or OpenAI releasing a new model. It could be something that increases the hype for these new model drops. But just recently, Open Router tweeted that they have onboarded a new stealth model. This is their second stealth model this past month, and it's a powerful foundational model that is optimized for general tasks, but especially exceptional at coding. It also features a 1 million token context window, just like Quazar Alpha, which is another model that dropped quietly last week. Now, here's where things get quite interesting. Nobody knows who built Optimus Alpha, which is the new model that just released. But the thing is, all signs point to the same mysterious creator behind the Quazer model, which actually released last week. In fact, there are reports when you actually prompt with the Optimus Alpha model directly, that it says that it's built by Chad GBT. And not to mention, Sam Ultimate actually replied to a post about it, hinting that it might be indeed OpenAI's new possible model, rumored to be the O4 Mini being tested out in the wild. Now, just like Quazer Alpha, this new Optimus Alpha model is blazing fast with inference speeds, and it ranks number two on the ELO leaderboard with a score of 1300, right behind Clawude 3. 7 Sonnet, which is a pretty impressive feat. And it also outperforms Gemini 2. 5 Pro in certain programming languages which is just wild. Its coding capabilities are next level where it handles complex programming tasks like multifile coordination and front-end development with precision comparable to cloud 3. 7 Sonnet. And on the creative writing v3 benchmark, it scored impressively, even edging out Gemini 2. 5 Pro in some creative tasks. But let's be real, it's obvious this model was built specifically for coding and technical work. There have been reports where they state that this model is exceptionally well at multifile coordination, superior aesthetic sense in design, intuitive understanding with complex systems, lightning fast response times, and lastly, you have the ability to switch between extraction levels. Before we get started, I just want to mention that you should definitely go ahead and subscribe to the world of AI newsletter. I'm constantly posting different newsletters on a weekly basis. So, this is where you can easily get up-to-date knowledge about what is happening in the AI space. So, definitely go ahead and subscribe as this is completely for free. Now, the best part about this model is that it's completely free and then you can actually access it as an API for free. As you can see on open router, they charge zero per 1 million input and 1 million output tokens. So you can go ahead and create an account and create an API key to access this Optimus alpha model so that you can deploy it in something like Klein or Root Code or many of the other AI coding agents. Now here's a test by a developer on Twitter and essentially he had wrote write a Python program that simulates realistically bouncing balls and other objects affected by gravity inside a square. Now this is the generation that we got from the quazar alpha model which was the first stealth model released on open rider and this is what we saw from Optimus Alpha. This is delivering a clean efficient code with advanced physics handling these different balls within this frame. And you can see that both of the models are exceptionally well at generating code but the Optimus model shows stronger system level reasoning whereas the Quazer model is uh better at speed and simplicity. So maybe this could be the new open AI04 model or it could be even the Gemini 2. 5 Flash. Now there is a prompt to test these different models in a sandbox game like Minecraft. And this is where it asks which AI generated this Minecraft build better where it's comparing the GPT4 Omni Mini versus the Optimus Alpha. And in this case, it was requested to build an iconic scene of people dismantling the Berlin Wall. And you can see that on the right hand side the Optimus alpha model was capable of actually generating that. Now I know this is a wild uh test but we're able to get a good understanding of how it's able to reason as well as showcase strong system performance in generation. Now I personally want to test this out with client. So what we're going to do is set this up with VS Code. If you do not know what client is, it's autonomous AI coding agent that you can directly integrate within any of these IDs like VS Code, Cursor, Windsurf as an extension. So go ahead and install it. It's completely free for your operating system. And then what you can do is open up VS Code and then you can install the extension through their extension store.

So you can search up the extension over here, client, download it, and then you will be able to access it on the lefth hand panel. Make sure it is on the latest update. Now once you have installed it, open up client and then you can go to the settings tab and you want to select the open router API provider. Paste in your open router key that you can generate completely for free after you create an account and then you want to search up the new Optimus model and you can then select the Optimus alpha model and then click done. Let's test out its front-end capabilities. This is where we're going to request it to create a landing page for a SAS website. And right away it is working on taking on this task. it is going to generate the plan for it and once it finishes generating the plan it'll rapidly go ahead and start generating the code. Now as you can see it is slightly delayed in terms of the generation and this is because there's so many people trying to generate with this model. If you actually take a look at the usage for this model, it is actually insane because in terms of the activity, we have record high activities where you have 38 billion tokens processed per day which is just insane and this is because a lot of people are finding out about this model due to its uh capability in coding. So let's see what it actually generates for this landing page. And there we go. We have this SAS landing page which actually looks super intuitive and it looks pretty modern. Now, if I am to compare the outputs of different landing pages, this is something that ranks pretty high in comparison to the other models. Gemini 2. 5 Pro does a slightly better job at this, but Claude 3. 7 also does I think it's at the best at the moment in terms of front-end development, but this model does a pretty great job. In my opinion, I would rank it number three in this category. Obviously, this is just one test I'm showcasing, but from what I've saw from previous generations, I am definitely stunned to see that it is pretty remarkable in generating front-end as well as complex code for these different services. But overall, what I think is a edge over the other models is that this model is completely free, guys. you have zero dollars being outputed for these types of generations, which is why I highly recommend that you try using it as much as possible because this is something that will help you prototype quite free and with great results. So that is why I'm making a video on Optimus Alpha. I'm going to be making another video showcasing its capabilities on the main channel. But that's basically a gist of today's video. This is something that is remarkable. The only thing that is kind of hindering it success is its max output. And another thing is that all prompts and completions for this model are logged by the provider and may be used to improve the model. Meaning that they will use whatever generation or code that you use or work with will be used as training. So just keep that in mind. I should have said this at the start, but this is kind of important if you're going to be working with this model for free. But that's basically it, guys. I hope you enjoyed today's video, got some sort of value. I'll leave all the links that I use in today's video in the description below. Make sure you subscribe to the newsletter, follow me on the Patreon, follow me on Twitter, and lastly, make sure you guys subscribe, turn on notification bell, like this video, and please take a look at our previous videos cuz there's going to be a lot more content posted here so that you can stay up to date with whatever is happening in the world of AI. But with that thought, guys, have an amazing day, spread positivity, and I'll see you guys really shortly. He suffers.