jevos
jevos is an open-source, self-hosted alternative to a hosted yes/no decision service called Jev: it answers a yes/no question about a piece of text on a laptop CPU in 50 to 220 milliseconds and speaks the same wire format so existing client code keeps working.
jevos answers yes/no questions about a piece of text on a laptop CPU, returning the probability that the answer is yes in 50 to 220 milliseconds. It is downloaded as a GGUF model file and served locally with a small command-line tool built on llama.cpp, with a context window of 8,192 tokens.
Its HTTP API speaks the same wire format as a hosted service called Jev, so code already written against Jev's SDK works unchanged for yes/no questions, though jevos does not yet support multiple-choice or scored questions. A request can bundle several named yes/no questions against the same input text, which is read once and answered for all of them together, and every response includes a timing header showing the inference time.