A legal team is building a Foundry-hosted contract-review tool that must read an entire 200-page merger agreement in a single request and produce a compliance summary. Early testing with a compact general-purpose model in the catalog causes the request to be rejected because the document exceeds the model's input limit. Which model catalog specification should the team check first when selecting a replacement model?

The model's maximum context window is the specification to check first, because it caps how many tokens, including input plus expected output, a model can process in a single call. The request failed because the 200-page merger agreement exceeded the compact model's input limit, so the team needs a model whose context window is large enough to hold the entire agreement text before it can attempt a compliance summary. The model's supported output format is not the first check, because output format controls how the response is structured, not how much input the model can accept, and changing to plain text would not fix a rejected oversized request. The model's content safety default severity settings are not the first check, because those settings govern whether flagged content is filtered or blocked, which is unrelated to a request being too large to process at all. The model's regional availability is not the first check, because regional availability determines where a model can be deployed for latency or compliance reasons, but it has no bearing on how much text a single request can contain; a model available in the right region could still reject the document if its context window is too small. Checking context window size first lets the team narrow the catalog to models capable of handling the full document.