xTuring provides a simple API for personalizing open-source LLMs through fine-tuning. It supports a range of models including LLaMA, GPT-OSS, Qwen, Mistral, and GPT-J, using parameter-efficient techniques like LoRA adapters and low-precision inference (INT8/INT4) to reduce compute costs. The library handles the full workflow from data preprocessing through training to inference and evaluation, enabling users to run private, personalized models locally or in their own cloud environment.