Anara offers some AI models that were developed by companies based in China, including Kimi, DeepSeek, and GLM. When you use these models in Anara, your documents and questions are processed on servers located in the United States. Nothing you upload or type is ever sent to China, or to the companies that created these models.

These companies released their models as open models. That means they published the model itself, a large file of numbers, for anyone in the world to download and run on their own computers. It works a bit like a recipe book: the authors published the recipes, but the cooking happens in your own kitchen. Anara runs these models on servers in the United States operated by US-based AI infrastructure companies, the same way we run our other models.

A model on its own cannot send information anywhere. It only produces answers on the computer where it runs. The companies that created these models have no connection to Anara's servers and no way to see your data.

We offer these models because they are among the best available for certain tasks and many researchers ask for them. If you'd rather not use them, every other model in Anara works the same way, and your data is handled identically on all of them.

We show this note because we think you should always know where your data is processed.