Follow-up from "Introduced chat capabilities to Explain Code"

The following discussion from !117674 (merged) should be addressed:

• @jerasmus started a discussion: (+2 comments)

  question (blocking): how did we determine this value? I think the value we use on the backend is 300 (ref).

  suggestion (non-blocking): WDYT of passing this value (and possibly the value for MODEL_MAX_TOKENS) down from the backend? Right now we'd have to maintain it on both frontend and backend.