# llama.cpp v b11810 adds new dequant switches and more binaries

> The open‑source LLM runtime released binaries for macOS, iOS and several Ubuntu builds, and merges fp16/fp32 dequant switches.

Oossa · 2026-10-22 · https://oossa.com/en/llama-cpp-v-b11810-adds-new-dequant-switches-and-more-binaries

ggml‑org released llama.cpp version b11810 on 22 Oct 2026. The update merges the fp16 and fp32 dequantisation switches, a change tracked as #30740. Pre‑built binaries are now available for macOS Apple Silicon, macOS Intel, iOS, and a range of Ubuntu CPUs, Vulkan and CUDA‑12.8 GPUs. The release page links to the tar.gz files for each platform.

## The facts

- Release tag: b11810 (published 2026‑10‑22).
- Binaries provided for macOS (arm64, x64), iOS, Ubuntu x64/arm64/s390x, Vulkan and CUDA‑12.8.

## Why it matters

Developers can run LLaMA models locally on more hardware without building from source.

## Sources & references

1. [ggml-org/llama.cpp b11810](https://github.com/ggml-org/llama.cpp/releases/tag/b11810) – llama.cpp, 2026-10-22

Last updated: 2026-10-22