# llama.cpp release adds GQA support and new tokenizer features

> The open‑source llama.cpp library now handles non‑standard GQA models and a BPE tokenizer, with expanded platform builds.

Oossa · 2026-10-22 · https://oossa.com/en/llama-cpp-release-adds-gqa-support-and-new-tokenizer-features

The ggml‑org team released version b11819 of llama.cpp on 2026-10-22. The update adds support for non‑standard GQA (grouped query attention) models and a BPE (byte‑pair‑encoding) tokenizer, plus a new domyn‑small pre‑type. It also adds metadata handling for Metaspace prepend schemes. The build matrix now lists many CPU and GPU targets, including Apple Silicon, CUDA 12/13, Vulkan, ROCm and Android Snapdragon.

## The facts

- Release date: 2026-10-22
- Adds GQA and BPE tokenizer support

## Why it matters

Developers can now run a wider range of LLMs, including those using GQA, on more devices such as smartphones and GPUs.

## Sources & references

1. [ggml-org/llama.cpp b11819](https://github.com/ggml-org/llama.cpp/releases/tag/b11819) – llama.cpp, 2026-10-22

Last updated: 2026-10-22