TypeSafe AI claims Jev hits 70ms to 500ms, 40x to 200x faster than frontier models

TypeSafe AI claims its early-access model Jev delivers 70ms to 500ms end-to-end responses for System One-shaped queries, 40x to 200x faster than frontier models. Jev uses parallel sampling to generate all outputs in a single query, rather than sequential token-by-token generation.