> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bytez.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Open vs Closed models

> We handle open & closed models differently

Our API simplifies working with a wide variety of AI models, including both popular closed-source options and flexible open-source alternatives. Although how we handle requests differs behind the scenes depending on the model type, you benefit significantly from our **Unified Model Protocol**.

This protocol means you can use the *same input format* to interact with *any* model on our platform—whether it's open or closed-source. It makes experimenting and switching between models much easier, almost like swapping Lego bricks in your project. This consistency frees you up to focus purely on building your application logic, rather than managing different provider interfaces.

Now, let's dive into the specific ways we handle requests under the hood for open vs. closed models to provide this seamless experience.

<AccordionGroup>
  <Accordion icon="lock-keyhole" title="Closed-Source Models (e.g., OpenAI, Anthropic, Gemini)">
    Think of us as a smart, multi-lingual translator and secure messenger when you use closed-source models. Our **Unified Model Protocol** means you use **one consistent format** for your requests and receive responses in **one consistent format**, regardless of the underlying provider.

    **The Process:**

    <Steps>
      <Step title="You Send Request">
        Your app sends an API request using our standardized input format
      </Step>

      <Step title="We Translate Input">
        We automatically translate your request into the specific format required by
        the chosen model provider (e.g., OpenAI, Google Gemini)
      </Step>

      <Step title="Forward Request">
        We securely pass your request to the model provider's API, using your API
        key, so the provider knows it's from you
      </Step>

      <Step title="Provider Computes">
        The provider runs inference on their servers
      </Step>

      <Step title="We Translate Output">
        We receive the provider's raw response and translate to standardized JSON
      </Step>

      <Step title="You Receive Response">
        Your app gets inference results back in standardized JSON
      </Step>
    </Steps>

    **Key Takeaway:** For closed-source models, we act as a router and standardization layer. You interact with a **single, unified protocol**, making it easy to switch between models providers or use multiple providers without changing your code structure. The inference itself happens on the provider's infrastructure.
  </Accordion>

  <Accordion icon="lock-keyhole-open" title="Open‑Source Models –  Serverless GPU Inference">
    When you run an **open‑source** model, Bytez handles all the heavy lifting for you.

    When you make a request to our API, this is what we do:

    <Steps>
      <Step title="Start a model container">If the model is not immediately available, we spin up a model container on our infrastructure.</Step>

      <Step title="Route requests">
        When the model is ready for inference, we route your request to the first available instance.
      </Step>

      <Step title="Scale according to load">
        As requests come in, we scale automatically, ensuring that your scaling demands can be met
        regardless of the model you choose.
      </Step>
    </Steps>

    All you need to worry about is specifying a model and making requests to the API, we take care of the rest!

    Our goal with open source models is to make them as easy and affordable to use closed source models.
  </Accordion>
</AccordionGroup>