Run Ling 3.1 Flash FREE in OpenCode

Julian Goldie (0:00)

Ling 3.1 Flash just dropped, and it's free right now inside OpenCode. Imagine giving your AI agents a brand new brain that thinks fast, remembers a ton, and works you around the clock for free. That's exactly what this is. Today, I'm going to show you how to grab it, how to set it up in a couple of clicks, and how to build a full website or app with it from one single prompt. Plus, I'll show you a powerful trick later in terms of how to drop this free brain into your favorite AI agent harnesses and make them way stronger. There's one thing you absolutely have to do first, so it won't even show up for you. SIGME and I'll show you exactly what it is. Let's get into it. So there's a brand new model from China called Ling 3.1 Flash, and it's now available for free on OpenCode. I'm going to show you exactly how to use it. It's 560 billion parameters CNC total, 25B active, so that means it's a mixture of experts model and 262K context window. And it's inclusion AI's latest model. Now, if you're wondering how it performs, we can check out the benchmarks on KiloAI as you can see right here. And again, this is free to use right now. You can also use it with KiloCode as well. So you can use this across multiple different places for free. You can see the weekly token usage. So it's grown in popularity right here. And this is just a good model that you can use. It's not going to be, for example, like Claude 5.5 level, but at the same time, it's free. And you can use this for sub-agents or for running your agents, as you can see right here. Now, if you're wondering how to use it, you can get it inside KiloCode. I actually use it inside OpenCode, so you can see OpenCode over here. And we have Ling 3.1 Flash Free ready to go. If you can't see it inside your list, you just go to check for updates inside OpenCode, make sure you're updated. If you don't know how to get OpenCode, you can just go to their website and get it for free. So from here, we can switch between build and plan mode, and then we can select Flash Free, and then we can switch as well between, do you want low, medium, or high thinking modes? I'm just going to go with the default for now. And if we say, for example, OK, build out a new beautiful website for the AI Profit boarding.com. And bear in mind as well, you can use this inside your harnesses. So for example, if you're running DeepSeaKarness, you could plug 3.1 Flash directly into DeepSeaKarness, which would be pretty nice too. And then what I'm going to do over here is we're just going to get the information from this page.

And I'm going to plug that into Ling 3.1 Flash here. So we'll paste that information. Make a more beautiful, interesting version of this.

Now you can also see here that we can add a project. So I'm just going to create a new project folder, as you can see right here, and we'll call this New Website for the AI Profit Boardroom. Click Create, open up, hit Enter, and then we just have to select an agent and model.

And you can see that's beginning to work right now. Now, in the meantime, whilst we're waiting for that to run, so you can see that running on the side over here. So you can see it's beginning to work now.

Bear in mind, sometimes the API on the free server from OpenCode doesn't work first time around, so it will just keep re-trying if that makes sense. I was actually looking at fixing this with Claude just a second ago. I just want to be 100% transparent with you, because if you're using it and you're like, why didn't it work first time around? This is why. So if you go back to the conversation with Claude, for example, you can see here that I actually gave it the screenshot of, why did it come up with that error first time around? So basically, that just comes from the free server. Sometimes it's briefly down for the link model, but it's back up. So within a couple of seconds, it was working. It actually took four attempts before it's finally working, and you can see it's now working again, which is great. And it's beginning to respond. Perfect. Now also, one of the best things about OpenCode is it has loads of free models. You can see right here. So this is inside OpenCode Zen. So for example, if we type in free here, you can actually see that there's loads of free models when you search. So for example, you've got link 3.1, Longcar, MIMO, MuSpark, Nematron, Union Alpha, etc. Some of those models will expire, but you can see all the different free models you can use, which is really good. And you can use Omniroot as well inside there, which is great as well. So loads of options, and you can see it's beginning to start working on this directly here too. Now you can also plug this into multiple different apps. So you can see here, for example, Inclusion AI. Let's see what it's being used on directly inside the app section. So you can see Claude Code is number one. So a lot of people are taking this model and then plugging it into a different body, if that makes sense. So I would say the brain is Link 3.1 Flash itself, but then you want to think of the harness. It's kind of like the body that takes the actions, right? So you've got Claude Code, Kilo Code, Hermes Agent as well.