Run AI models locally on your machine with node.js bindings for llama.cpp. Enforce a JSON schema on the model output on the generation level
node-llama-cpp.withcat.ai
withcatai/node-llama-cpp