Run LLMs on Your Phone — On-Device AI

Browse iPhones and Android phones for running quantized LLMs on-device. See which models fit in 8–16 GB of unified memory and run without an internet connection.

On-Device LLM Inference on Smartphones

Modern smartphones can run small quantized LLMs entirely on-device with no internet connection and no cloud costs. iPhones with A17 Pro or M-series chips and high-end Android devices with Snapdragon 8 Gen 3 are best suited for 1B–7B parameter models at aggressive quantization (Q4 or lower).