Joey-1 returns a calibrated probability instead of prose, costs a cent per million tokens and ships as open…
Microsoft and Nace both shipped small decision-scoring models that rank choices rather than generate prose.
TypeSafe, maker of the non-text Jev model, has raised $870M at a $7.5B valuation led by Andreessen Horowitz…
Strands Decider 2B returns calibrated choices and scores instead of generated prose, running entirely on local hardware.
Cloudflare's first in-house models, Clef and Clef-flash, return probabilities instead of generated text.
d1 returns calibrated probabilities for fixed choices and spends zero output tokens on the answer.
Nokia's applied research team open-sourced AnyJev, a Python library that turns any open model into a decision maker…