Local Ai
b9161
Release b9161 of llama.cpp includes enhanced regex handling for Qwen3.5 tokenizer, adding a custom unicode handler to prevent stack overflows on long inputs . The release also adds SYCL Level Zero SDK
Release b9161 of llama.cpp includes enhanced regex handling for Qwen3.5 tokenizer, adding a custom unicode handler to prevent stack overflows on long inputs . The release also adds SYCL Level Zero SDK support with compile-time and runtime flags to separate Level Zero and SYCL memory API paths .
Source: llama.cpp Releases | 2026-05-15