@royinblr11 

You're absolutely right to question the answer — the correct model for an application needing a foundation LLM with a large context window is: DBRX

Why DBRX is the Best Fit:

• It is a foundation model, designed for generation tasks.
• It supports a 32K token context window out of the box — ideal for handling long documents, chats, or code.
• It is open-source and production-ready for enterprise-level generative AI tasks.

So, if the goal is:

Build a generative application using an open-source model with a large context window,

Then the best answer is DBRX, not MPT-30B.