Local LLM
A local LLM is a large language model that runs on a person's own computer, through a runtime such as Ollama, LM Studio, or llama.cpp, instead of being called from a cloud AI provider.
A local LLM is a large language model that runs on a person's own computer, through a runtime such as Ollama, LM Studio, or llama.cpp, instead of being called from a cloud AI provider.
Network tools mostly miss it. Once the model file is downloaded, a local LLM answers without calling the internet, so DNS, proxy, and CASB logs see the download at most. Prompts stay on the machine, which lowers the risk of data leaving the company, but local models bring their own questions: unvetted model files and licenses, local AI servers exposed to the network, and outdated runtimes. That makes them a form of shadow AI that only shows up on the endpoint (see endpoint AI visibility).
Behest Radar recognizes local models run with Ollama, LM Studio, or llama.cpp, and names the model. For the processes, ports, model files, and fleet checks to look for yourself, see how to find local LLMs on employee laptops.