How to Access and Use the Free GLM 5.2 AI Coding Model with NVIDIA Cloud Code

Artificial intelligence coding tools are becoming increasingly expensive, with many developers paying hundreds of dollars monthly for access to powerful models. However, there is a new, completely free AI coding model called GLM 5.2, accessible through NVIDIA, that offers a competitive alternative. This article will guide you through connecting GLM 5.2 to cloud code using NVIDIA's free API, setting up the environment, and testing the model on a real project.

GLM 5.2 is one of the latest open AI coding models and is gaining popularity for its performance. It competes with premium models like OpenAI's GPT-4.8. The best part is that GLM 5.2 is free to use and can be accessed via NVIDIA's platform.

Accessing GLM 5.2 on NVIDIA's Platform

1. Open your browser and navigate to build.nvidia.com.
2. On the homepage, look for GLM 5.2 at the top. If you don't see it immediately, click on the "Models" tab.
3. On the Models page, you can find GLM 5.2 listed among the popular models. Alternatively, use the search bar to type "GLM" and press enter.
4. Click on GLM 5.2 to open its page. Here, you will find information about the free endpoint and downloadable options.

Setting Up NVIDIA Cloud Code

To use GLM 5.2 with cloud code, you need to install NVIDIA Cloud Code:

1. Search for "Cloud Code install" to find the quick start documentation.
2. Follow the single command provided to install Cloud Code on your machine (compatible with Windows, Mac, and Linux).
3. After installation, open Cloud Code. You will be prompted to log in; skip this step for now.

Integrating GLM 5.2 with Cloud Code Using Jan.ai Proxy

NVIDIA's APIs are OpenAI compatible, but Cloud Code requires Anthropic-compatible APIs. To bridge this gap, use a third-party proxy called Jan.ai:

1. Download Jan.ai from its official site (available for Windows, Mac, and Linux).
2. Install and open the Jan.ai app.
3. Navigate to the settings within Jan.ai.
4. In the providers list, find and enable the NVIDIA NIM provider.
5. You will need two pieces of information:
   • API Key
   • Base URL

Obtaining the API Key and Base URL

1. Go back to the NVIDIA website and log in or create an account if you haven't already.
2. Verify your account via phone number.
3. After logging in, navigate to the API keys section.
4. Delete any old keys and generate a new API key with a 1-hour expiry.
5. Copy the generated API key.
6. The base URL is https://integrate.api.nvidia.com/version1.

Configuring Jan.ai

1. Paste the copied API key into Jan.ai's NVIDIA NIM provider settings.
2. Ensure the base URL is correctly set.
3. Refresh the models list in Jan.ai; you should see six new models added, including GLM 5.2.

Configuring Cloud Code to Use GLM 5.2

1. In Cloud Code, go to the left sidebar and select the Cloud Code section.
2. Change the default model to GLM 5.2 by selecting it from the NVIDIA models list.
3. Optionally, change the secondary model to KMK 2.6.
4. Save and enable the configuration.
5. Cloud Code will prepare the server and notify you when the environment variables are updated.
6. Relaunch Cloud Code in a new terminal to apply changes.

Testing GLM 5.2 with a Real Project

1. Open your project directory in a terminal.
2. Run the cloud command to launch the Cloud Code CLI.
3. Confirm that GLM 5.2 is the selected model.
4. Start interacting with the model by typing queries.

Example Interaction

• Ask the model to describe the project in the current directory.
• The model will explore the project files and provide a detailed summary.
• In the test, the model quickly identified the project as a TypeScript and Tailwind CSS application, explaining its structure and key files.

The response time was impressively fast, around 9 seconds, demonstrating the model's efficiency even under heavy usage.

Conclusion

GLM 5.2, accessible for free through NVIDIA's platform and integrated with Cloud Code via Jan.ai, offers a powerful AI coding assistant without the high costs associated with premium models. Its performance rivals top-tier models, making it an excellent choice for developers seeking cost-effective AI tools.

If you found this guide helpful, consider exploring more free AI tools and sharing your experiences to help others benefit from these resources.