← Blog

2026-10-26 · 5 min read

How to Effectively Use Air Credits to Access AI Models

Discover how developers can enable models, purchase Air Credits, and utilize the /api/v1 chat completions endpoint in the Air Inference marketplace.

How to Effectively Use Air Credits to Access AI Models

In the rapidly evolving landscape of artificial intelligence, developers are continuously seeking efficient ways to access powerful AI models. Air Inference has emerged as a robust two-sided marketplace that connects developers with providers offering OpenAI-compatible endpoints. By leveraging Air Credits, developers can easily enable models, make API calls, and integrate advanced AI capabilities into their applications. This article will guide you through the process of purchasing Air Credits, enabling models, and effectively utilizing the /api/v1 chat completions endpoint.

Understanding Air Inference

Air Inference serves as a bridge between developers looking for AI capabilities and providers who offer various AI models. It operates on a simple premise: developers purchase Air Credits which they can use to access different AI models listed by providers. This two-sided marketplace not only simplifies the process of accessing AI services but also ensures that providers can monetize their endpoints efficiently.

What are Air Credits?

Air Credits are the currency used within the Air Inference platform. Developers top up their accounts with these credits, which can then be used to call APIs for various AI models. Each API call deducts a certain number of Air Credits based on the complexity of the request and the model being accessed. This model allows developers to pay only for what they use, making it a cost-effective solution for integrating AI into applications.

Step 1: Purchasing Air Credits

Before you can start accessing AI models, you need to purchase Air Credits. The process is straightforward:

  1. Create an Account: Visit the Air Inference website and sign up for an account. This will require you to provide some basic information and verify your email address.

  2. Top-Up Your Account: Once your account is created, navigate to the account settings or billing section. Here, you will find options to purchase Air Credits. The platform typically offers various packages, allowing you to choose the amount that fits your needs.

  3. Payment Method: Select your preferred payment method to complete the purchase. The platform may accept several forms of payment, ensuring flexibility for users.

  4. Check Your Balance: After confirming your purchase, check your Air Credits balance in your account dashboard. This balance will be used for future API calls to the models listed on the platform.

Step 2: Enabling Models

With your Air Credits in place, the next step is to enable the AI models you wish to use. Air Inference supports a variety of OpenAI-compatible endpoints, including vLLM, llama.cpp wrappers, and RunPod among others. Here’s how you can enable models:

  1. Browse Available Models: Access the marketplace section of Air Inference to view the list of available AI models. Each model will have a brief description, performance metrics, and the required Air Credits for calling the API.

  2. Select a Model: Choose a model that aligns with your project requirements. Consider factors such as the model's capabilities, latency, and the cost in Air Credits per API call.

  3. Enable the Model: Once you have selected a model, click on the "Enable" or "Activate" button. This action will prepare the model for your API calls. Some models may require you to agree to specific terms or usage policies before enabling them.

  4. Documentation Review: After enabling the model, it's crucial to review the documentation provided by the provider. This will give you insights into the API's capabilities, required parameters, and response formats.

Step 3: Calling the /api/v1 Chat Completions Endpoint

Now that you have your Air Credits and have enabled your desired models, it’s time to make API calls. The /api/v1 chat completions endpoint is one of the primary means through which developers can interact with these AI models. Here’s a step-by-step guide on how to call this endpoint effectively:

  1. Set Up Your Development Environment: Ensure you have a development environment set up where you can write and test your API calls. You can use tools like Postman, Curl, or any programming language that supports HTTP requests.

  2. Construct Your API Request: The structure of your API request will depend on the specific model you are using. Generally, you will need to include the following elements:

    • Endpoint URL: This will typically be https://api.airinference.com/api/v1/chat/completions.
    • HTTP Method: Most calls to this endpoint will use the POST method.
    • Headers: You will need to include authentication headers, typically in the form of an API key, along with content type headers.
    • Body: The body of your request will contain the input for the model, such as the prompt for a chat completion.
  3. Example API Call: Here’s a simple example of what a curl command might look like for calling the /api/v1 chat completions endpoint:

    curl -X POST https://api.airinference.com/api/v1/chat/completions \
    -H "Authorization: Bearer YOUR_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
        "model": "desired-model",
        "messages": [{"role": "user", "content": "Hello, how are you?"}]
    }'
    

    Replace YOUR_API_KEY with your actual API key and desired-model with the model you have enabled.

  4. Handle the Response: After making the API call, you'll receive a response from the model. This response will typically include the generated text based on your input. Make sure to handle any potential errors, such as insufficient Air Credits, invalid API calls, or issues with the model itself.

  5. Monitor Your Usage: Keep track of your Air Credits balance as you make API calls. The Air Inference dashboard provides insights into your usage, allowing you to manage your credits effectively.

Best Practices for Using Air Inference

To maximize your experience with Air Inference, consider the following best practices:

  1. Start Small: If you’re new to using AI models, start with smaller, less complex tasks to familiarize yourself with the API and its capabilities. This approach will help you understand how to structure your requests and interpret responses effectively.

  2. Optimize Your Queries: Crafting effective prompts is crucial for getting the best results from AI models. Experiment with different phrasings and structures to see how they affect the model's output.

  3. Leverage Documentation: Always refer to the documentation provided by the model's provider. It contains valuable information about usage limits, optimal parameters, and examples that can help you refine your calls.

  4. Stay Updated: Keep an eye on any updates or changes to the models you are using. Providers may release new features or improvements that can enhance your experience.

  5. Participate in the Community: Engage with the Air Inference community through forums or social media platforms. Sharing experiences and learning from others can provide insights that you might not find in documentation.

Conclusion

Air Inference offers developers a streamlined way to access powerful AI models through the use of Air Credits. By following the steps outlined in this article—purchasing credits, enabling models, and effectively calling the /api/v1 chat completions endpoint—you can integrate advanced AI capabilities into your applications with ease.

As the marketplace continues to grow, developers can look forward to a wide array of AI models, enabling innovative applications across various industries. Embrace the power of AI with Air Inference and start exploring the possibilities today.