A model runner is software that loads an AI model and runs it on your computer. It handles the processing needed to turn inputs, such as prompts, into outputs, such as generated text.
For local language models, the runner manages tasks such as loading model files, allocating memory, and using supported CPU or GPU hardware. It may also provide controls for generation settings and a local API that other applications can connect to.
A model runner and a user interface serve different purposes. The runner executes the model, while the interface provides a way to interact with it. Some applications combine both; others let you connect a separate interface to the runner.
When choosing a runner, check its supported model formats, hardware requirements, and connection options. Start with one that fits your system and your first workflow, then expand as your needs grow.
