TypeSafe AI launched a set of onboarding tools, including a primer and a custom interactive environment, to demonstrate how its model handles typed questions. The tool is now available for immediate use after the company removed a previous access waitlist, allowing any user to explore specific use cases and functional constraints via a digital playground.

The Jev model provides typed answers—such as probabilities, a chosen option, or a rubric score—instead of generating standard text. It is currently offered in beta on the Venice API with a 64K token limit per call and a cost of $0.042 per 1M input tokens, while output tokens are free.