Claude Code for FREE Forever — OpenRouter Setup (2026)

Kurze Zusammenfassung

Dieses Video zeigt, wie man Cloud Code kostenlos mit OpenRouter nutzen kann, ohne die Einschränkungen lokaler Modelle oder hohe Kosten. Es werden kostenlose Modelle und die Einrichtung einer API erläutert, um Cloud Code ohne monatliche Gebühren oder Ratenbeschränkungen zu verwenden.

  • Nutzung von OpenRouter für kostenlose AI-Modelle.
  • Konfiguration von Cloud Code zur Nutzung von OpenRouter.
  • Vermeidung von Kosten und Ratenbeschränkungen.

Das Kostenproblem mit Claude Code

Die Nutzung von Cloud Code Pro kann monatlich etwa 20 US-Dollar kosten, wobei die Kosten bei starker Nutzung auf 50 bis 200 US-Dollar steigen können. Zudem gibt es Ratenbeschränkungen. Um diese Kosten zu vermeiden, wird OpenRouter vorgestellt, das eine kostenlose Alternative bietet. OpenRouter ermöglicht den Zugriff auf verschiedene kostenlose Modelle über eine API, wodurch die Notwendigkeit entfällt, für Cloud Code zu bezahlen.

Was ist OpenRouter?

OpenRouter ist ein einheitliches API-Gateway, das Anfragen an über 200 KI-Modelle weiterleitet, darunter solche von Cloud, Gemini, DeepSea, Qwen, Lama oder NVIDIA. Einige dieser Modelle sind kostenlos nutzbar. Es ist keine Kreditkarte erforderlich, sondern lediglich ein API-Schlüssel.

Beste kostenlose Modelle für Coding

Empfehlenswerte kostenlose Modelle sind OpenRouterFree, Step 1, 3.5 flash, NVIDIA, Metro, Nemetron, RCAI, Liquid und LFM. OpenRouterFree ist besonders nützlich, da es Anfragen automatisch an verfügbare kostenlose Modelle weiterleitet, ohne dass spezifische Konfigurationen erforderlich sind.

Schritt 1 — Installiere Claude Code

Falls Cloud Code noch nicht installiert ist, kann es über NPM oder Curl installiert werden. Es wird empfohlen, die aktuelle Version von Cloud Code zu überprüfen. Unter Windows kann Cloud Code über WSL oder direkt unter Windows installiert werden.

Schritt 2 — Holen Sie sich Ihren kostenlosen API-Schlüssel

Um einen kostenlosen API-Schlüssel für OpenRouter zu erhalten, muss man sich zunächst ein kostenloses Konto erstellen. Anschließend kann man einen API-Schlüssel generieren, indem man auf "Create" klickt und einen Namen vergibt. Optional kann ein Ablaufdatum festgelegt werden. Der generierte API-Schlüssel beginnt immer mit "SK".

Schritt 3 — Konfigurieren & Ausführen

Die Konfiguration erfolgt über die settings.json-Datei im .cloud-Ordner des Projekts. Hier wird Anthropic so konfiguriert, dass Anfragen an die OpenRouter-API weitergeleitet werden. Der OpenRouter-API-Schlüssel wird in der entsprechenden Zeile eingetragen, während die Zeile für den Anthropic-API-Schlüssel leer bleibt. Das gewünschte Modell, z.B. OpenRouterFree, wird ebenfalls in dieser Datei definiert.

Live-Demo + Alias-Trick

Durch die Konfiguration der settings.json-Datei im Projektordner wird OpenRouterFree als Standardmodell festgelegt. Eine Live-Demo zeigt, wie schnell OpenRouter Anfragen beantwortet, ohne dass lokale Ressourcen stark beansprucht werden. Diese Methode ist besonders nützlich, um Cloud Code kostenlos zu nutzen.

Wichtige Erkenntnisse

Die Nutzung von OpenRouter ermöglicht es, Cloud Code kostenlos und ohne die Einschränkungen lokaler Modelle zu verwenden. Die Konfiguration ist einfach und erfordert nur wenige Änderungen in der settings.json-Datei. Dies macht Cloud Code zu einem funktionalen KI-Assistenten ohne zusätzliche Kosten.

Fasse jedes YouTube-Video in Sekunden zusammen

Hol dir Zusammenfassungen wie diese für jedes Video, in deiner Sprache, mit der kostenlosen Summ-App.

Laden im App Store
Jetzt bei Google Play

Transkript

Vollständiges Transkript von „Claude Code for FREE Forever — OpenRouter Setup (2026)“. Klicke auf einen Zeitstempel, um das Video an dieser Stelle zu öffnen. Du brauchst ein anderes Video? Nutze den YouTube-Transkript-Generator.

0:00 Cloud Code is one of the best AI coding assistant out there, but nobody talks about the facts that you can run it completely free forever without hitting the entropic rate limit or paying like $20 a month. And I'm going to show you exactly how to do this in under five minutes using open router. I have created a video a couple of days before how to run a cloud code completely free using OLAMA. But the challenge was like, you know, you need to have download the model locally and which consumes a lot of memory and, you know, it have a limitation that you cannot run a bigger model in your local machines because of the memory and the context it needed. In this tutorial, I'm going to show you

0:47 something. You don't even need to download anything into your local. It's going to be completely, you know, to the Internet and you're not running anything from your laptop. We're going to use Open Router and we are going to use some free models. There are different models where Open Router gives and we just need to create an API and we need to call these models to your cloud code. So here's the issue, right? When you use Cloud Code Pro, it's like $20 a month. And even like if you have heavy API usage, it hits like $50 or $200 per month. Or you have some rate limit that you cannot use more than when you are just paying $20. And what we are going to do is like we are going to use OpenRouter, which is 100% free. And there are a lot of models

1:36 which have completely free, which you can just connect using those API keys from an open router. So open router is like a unified API gateway that routes your requests like over 200 AI models whether it could be from cloud, Gemini, DeepSea, Qwenn, Lama or NVIDIA. There are different models which you can find from the open router and we just need to use the API to to get into one of those. So some of those models are completely free. There are also like paid one, but we will use the free one and there is no billing, no credit card, just an API key. And here are some of the free model which we'll be using or I can recommend for you to use. And

2:23 one of the best one or optional one is like OpenRouterFree which is like a set of model which you can automatically route because once you set this model OpenRouterFree it can route it automatically to the request based on the free models available. And you don't need to set any specific configuration. And you know there are like Step 1, you know 3.5 flash which is like like 256 context. You also have like NVIDIA, Metro, Nemetron, which is also good. You also have RCAI, Liquid, LFM. So there are different models which you can see, which are completely free. If you don't have Cloud Code installed on your machine, just run the NPM install or there is also the new installation using curl. I think the NPM is

3:10 more deprecated. So you can install Cloud Code and just check your Cloud Code version. And if you're using Windows, you can use a WSL or you can install it directly on the Windows as well. What we are going to do, we are going to set up a free API key for OpenRouter and we are going to use that. So as I said, OpenRouter, get an account with that, it's completely free. Then you go to, I already have an API key, but you can just click on create, give a name, I will give test then you can set an expiration or if you don't want to set an expiration just click click clear it it will give you an API key just copy this API key which always start with SK or right so I will delete this because I already have one so

3:59 once you have the API key the next step is super simple go to your cloud setting settings.json. You can do it in two ways. You can do it specific to your project. In my case, you know, I'm inside the project and I have the folder.cloud and you put this setting.json. And what you can do is then you can set this configuration. What we are basically doing is we are setting Anthropic to route it to the open router API. And we need to set your open API, the open router API key here, which we just created. Put it here and then keep this one always blank. You know, and drop it API key. Keep this blank. It is specifically done. Don't remove that line. Just keep it blank. Then you define

4:47 your model. So you can set the open router free or you can also set other ones. if you want to see some models you know you can go to the models in open router and you can see which model is best so you can see this one have different ones which it automatically route you can also search for free here and then it will list out you know different models so let me go to here you can search for free then it will list out some free models here or other option is you can also come to the categories or you know the prompt pricing you can just make it zero then it will list down you can see the nvidia lama nematron is completely free you can see zero million zero dollar per million tokens and zero dollar for output tokens

5:35 so you can simply use this one you know this would be you know the model number if you want to use there are more here so you can look at what is available then you can use one of those. So, basically I would like to use set for the open router free one. Then once you set it, you know, what you can do is this is my cloud. Okay, you can see it's using Sonnet and this setting is like a default one which I'm using from my home. But in this case, I logged in with... If I exit from here, if I log in again with cloud here you can see I'm using open router free because I'm inside this project and I

6:21 have my dot cloud and also my setting dot JSON if you put the setting dot JSON inside you know your home directory because by default cloud code have a no setting dot JSON inside your home directory then it will apply for every cloud cloud code instance but as for this demo i'm only setting it to the project and then you can see if i go to model you can see i'm using open router free right then you can simply say you know uh whatever you want then it will start you know you can see it's asking what do you want to do let me see what is uh open router let's ask You can see it given the output like OpenRouter is API. It's super fast, you

7:08 know, not like your local LLM, which you have to download and you need to have high CPU and high memory for running it. This is quite fast and you don't need to set up anything locally on this machine. So, I would tell you that, you know, use this method to set up your cloud code if you want to run something free and you don't have to pay for your you know, your cloud code, your instance, right? So, this would be a super helpful setup and I will link all these things to the video description. So, as I said, you know, I had this previous tutorial of Olama downloading this cloud code, but as I was telling, you know, it could be not that suitable because of the size

7:55 and the memory you need, you know, for running it. But this is super great, you know, this is completely simple work, just few lines of code in your settings.json and you can start using your cloud code completely free. So this is the configuration file which I was saying, you know, put this entropic settings in into the settings.json, that's it. Then you have zero model. You can also run with the hyphen hyphen model once you set up the router API, but I would prefer you set up your model in the Anthropic model by default. Then you don't have to make any changes, you can just set it up there. Super great, right? I hope you like this video and you can be a completely functional AI assistant at

8:40 zero cost. I'm sure you like this, you try this one out and drop a comment if you're already using CloudCode or if this is your first time. Let me know if this video helps you and if you need any help on setting up, I'm always ready to help you. Just put a comment into the video comment section if you face some problems. The full instructions, as I was showing, it's available in the video, I will put it into the video description. Just copy paste the template and just put it to your settings.json. Subscribe to my channel for more videos and if you want to see more videos like DevOps or DevSecOps, subscribe to my channel. Also like this video so that it can reach out to more people and

9:26 I will see you in the next video.

Share

Weitere Video-Zusammenfassungen