How to Run DeepSeek V4 Flash for FREE!

**Julian Goldie** (0:00)
The new version of DeepSeek, DeepSeek V4 Flash 731 is now available for free. I want to show you exactly how to use it. So what you can use is open code. Open code have just announced the DeepSeek V4 Flash. The new version is now available on the free tier inside open code, which means that you get the massively improved version of DeepSeek for free, and you can plug it into an agent operating system. Like you can see right here to use DeepSeek for free. So what we actually have is we have DeepSeek V4 Flash working with V4 Flash is the new version of DeepSeek. It's been updated inside open code so that we can build and automate anything we want. So if we plug in a prompt like this, for example, we can start using it for free.
And the other benefit of this is that you can fully customize it. So for example, if we look at our workspace over here, we've been building out with DeepSeek 731 inside our operating system, and it all gets saved inside our workspace, right? And you can actually use this to build beautiful websites and landing pages and all sorts of cool stuff. So here's an example of a website we created. You can see all this cool stuff that we've created over here. And we've got the system working with open code in the background using DeepSeek V4 Flash for free over here as well. So you can see it's now building out the app that we just asked it to create. And if we open that up, boom, shakalaka, we just created a app in the space of a couple of minutes using this free system, right? The other cool thing about this is if we go over to our workspace, anything that we build with DeepSeek V4 Flash, the new version is available right there, and it's saved to come back to later. Now, if you want to get this fully customized version of OpenCode, it's inside the AR Profile Boardroom, link in the comments to the arprofileboardroom.com. But the point here is this is exactly how you can run DeepSeek V4 Flash for free. Now, also we have Hermes Agent over here, and you can also plug in OpenCode into Hermes Agent 2 So there's actually a skill for this, it's called OpenCode with Hermes, and what that means essentially is you can use the Hermes OpenCode skill, so that every time you build something, you can code with DeepSeek V4 Flash for free.
So you can see the skill here, and basically this allows Hermes to delegate coding to OpenCode CLI. Why is that useful? Because OpenCode CLI has DeepSeek inside it. And so this is a skill that you can install directly to Hermes Agent, and then it can start installing this directly in there as well. So it's called OpenCode CLI, that is the skill to use directly with Hermes Agent. And if you wondered how to do this, you can literally just like copy and paste that into Hermes and say, listen, install this skill, and then just make sure you have OpenCode set up. If you don't have OpenCode set up, you can just go to the website and install it. Like it's pretty simple. You just go to opencode.ai.
Now, if we have a look here as well, the interesting thing about DeepSeq itself is that it's massively improved version versus, right? So the bigger, larger, more supposedly powerful model has now been leapfrogged by a small model with exactly the same size and architecture, but it's just been trained differently and it performs much better. So for example, if we actually go to Goldie Bench, which is where I test all this sort of stuff, we tested out DeepSeq for Flash, right? And the newest version is really, really good so far. And this is not like, you know, Fable 5 level, but it can create awesome stuff.
You can build out for free and it improved massively in the latest release. It just dropped a couple of days ago. Bear in mind as well, DeepSeq are going to be releasing DeepSeq Pro, another improved version of it, because both of these models were previously in preview. And so who knows what's going to happen next? Like I think it's only going to get better. And you know, you could be seeing DeepSeq Pro as like a Fable 5 level competitor or even overtaking it. And if that happens, that is going to be, that could change the whole industry essentially, because bear in mind, DeepSeq is open source, which means it's open weights. Whereas for example, Claude, OpenAI, et cetera, these are not like open source. You can't get the open weights for these models. So it can build some pretty cool stuff. It's easy to use. It's free. You can plug it into your agents like Hermes. You can use it with open code. And what I like to do personally, is I also have something called DeepSeq Coder, right? And so this is a fully dedicated coding window that I can use with DeepSeq. So for example, if I say over here, create a beautiful website for an SEO agency, we can then click build here. And once it finishes building, we can preview it on the right hand side. So we have the chat on the left hand side, we have the right hand side with the preview, you can see the code here.