How “Open Source” LLM Models Actually Make Money

The AI ecosystem has exploded since the release of ChatGPT, and fortunately, the open-source community around AI and LLMs has blossomed right alongside it. These open models have achieved groundbreaking performance, often competing head-to-head in benchmarks with leading commercial models from giants like Anthropic and OpenAI.

However, achieving this level of power is a monumental and expensive undertaking. It costs the organizations behind them—be they corporations or even entire nations—a massive amount of capital, ranging from millions to billions of dollars. You've likely heard of the big names: Llama, Qwen, DeepSeek, and many more.

This reality raises two important questions:

First, since most of these organizations are for-profit institutions, how exactly do they make money on models they seemingly give away?

Second, just how “open” are these open-source models, really? Releasing the model weights is only part of the story. In the worst cases, you might be deploying a so-called “open-source” model that isn't even compliant because it has biases in the training data you are not even aware of.