Long context

By default Jev keeps 64k tokens per request and 32k per question, a cap TypeSafe says it chose deliberately because all models, Jev included, are less reliable on long inputs. Sending TypeSafe-Danger-Zone-Long-Context: true raises both caps to 262k tokens on jev-1.13.0 and jev-1.14.0.

The release post is explicit that the longer window is a permission, not a recommendation: it strongly recommends limiting context length, and the docs repeat that long context costs the same per token but is slower and less accurate than the defaults. For agents, the shorter cap is a design constraint that pushes context management back into code.