Can Someone Explain Azure AI Foundry Pricing?

If you're just using AI Foundry to deploy OpenAI's large language models, you'll primarily pay per token. However, Azure AI Foundry has various resources, each with its own pricing model. While most services are consumption-based, some require you to manage the compute resources, for which you pay per hour. Services like document intelligence charge based on the number of pages or words processed. Keep in mind that Foundry simplifies deployment by automatically setting up multiple resources, but if you're self-funding, it's wise to keep an eye on your expenses. The monitoring dashboard can help you track your spending effectively.

Just to clarify, AI Foundry is geared towards self-hosting models, so that might also influence how costs work depending on what you're trying to deploy.

Just chiming in here because I'm new to this service too and trying to grasp its pricing model. Hoping to see some useful answers soon!

You can check out this pricing guide I found: https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/final/en-us/microsoft-product-and-services/azure/pdf/ms-azure-ai-foundry-pricing-guide-e-book-final.pdf. But I still have trouble finding the pricing details for models like Phi-4, Grok, and Mistral; it seems like they only list prices for OpenAI models.