Processing Your AI Data in Europe Can Cost More

Where your AI processes data can now affect the price. Mistral charges 10% more when customers choose regional AI processing in Europe or the United States. OpenAI also applies a 10% extra charge to eligible newer models when regional processing is used. For European businesses, this provides more control over where AI processing takes place. This means companies are increasingly choosing not only which AI model to use, but also where they want their AI processing to happen.

There are also ways to reduce AI costs. With Mistral, cached tokens can be 90% cheaper than normal input tokens, while batch processing can offer a 50% discount for work that does not need an immediate answer. OpenAI also offers lower cached-input and batch prices on supported models. AI pricing is therefore becoming more than one number: the model, processing location, caching and processing method can all change the final cost.