ggml-org releases llama.cpp v1.0.0

The release notes say more detail will follow at 3PM CET, and nightly builds such as b12359 continue alongside the tagged version. The release is downloadable from the project's GitHub releases page.

The changelog since v0.9.0 spans all major backends — CUDA, Vulkan, SYCL, Metal, OpenCL, WebGPU and CPU — plus server changes including SSRF and DNS-rebinding protections, opt-in tools for non-loopback addresses, and Qwen3.5 tensor merging during conversion. Obsolete examples were removed.