TypeSafe AI releases Jev, a model that outputs probabilities over fixed answer options
Overview
TypeSafe AI, a startup that came out of stealth on September 15, released Jev, a model that does not generate free-form text like LLMs.
Instead, it takes yes/no, multiple-choice, or rating questions with a fixed set of answers and outputs the estimated probability that each option is correct.
According to Understanding AI's Timothy B. Lee, this design lets the model be served faster and more cheaply than LLMs and fits ordinary if-statement logic. Lee says he used Jev to flag spam comments on his blog in place of Gemini 3 Flash; these are his own claims from a single use case.
Written by AI from the articles below · updated Oct 8, 7:58 PM ET
Check the sources:
Article timeline
Follow the coverage from different perspectives. Times are ET.
- Ruan Yifeng · Tech WeeklyWeekly tech digest examines Jev decision model, which returns probabilities instead of text
AITypeSafe AI released Jev, a "decision model" that returns a floating-point probability rather than text, which can answer yes/no and multiple-choice questions and score content against criteria. The source cites two browser-extension examples: semantic Ctrl+F search and webpage quality scoring. Simon Willison's criticism is that Jev offers no explanation for its numbers.
- Understanding AI (Timothy B. Lee)TypeSafe AI's Jev returns probabilities over fixed answers instead of text
AITypeSafe AI released Jev, a model that answers yes/no, multiple-choice, or rating questions by outputting the estimated probability of each option. The author notes this design lets the model be served faster and more cheaply than LLMs and fits ordinary if-statement logic, and says he used it to flag spam comments on his blog in place of Gemini 3 Flash.
Heat trend
Not enough continuous observations to show a trend yet.