https://github.com/Thireus/llama.cpp
this guy has support for qwen3 vl up on his tr/qwen3-vl-3 branch for anyone who wants to try it, worked for me with https://huggingface.co/yairpatch/Qwen3-VL-30B-A3B-Instruct-GGUF
prebuilt versions https://github.com/Thireus/llama.cpp/releases/tag/tr-qwen3-vl-3-b6981-ab45b1a
apparently there's more work that needs to be done to clean up the implementation for a PR since it breaks qwen2.5 VL or something but hopefully one is created soon