# Multi-scenario recommendation

Qwen3-14B-abliterated is a quantized version based on the Qwen3-14B model, optimized using llama.cpp, offering multiple quantization options to meet different performance requirements.

A quantized version based on the Steelskull/L3.3-MS-Nevoria-70b model, using llama.cpp for imatrix quantization, supporting multiple quantization levels for different hardware environments.