# llama.cpp adds Q1_0 and Q2_0 quant support for Hexagon NPU

> The open‑source LLM runtime now runs more efficiently on Snapdragon chips with new low‑bit formats.

Oossa · 2026-10-17 · https://oossa.com/en/llama-cpp-adds-q1-0-and-q2-0-quant-support-for-hexagon-npu

The llama.cpp project released a new version that adds Q1_0 and Q2_0 quantization for Qualcomm's Hexagon NPU. The update also tweaks tiling and vector dot operations to cut overhead. It works on Linux and Windows builds that target Snapdragon CPUs, Adreno GPUs and Hexagon NPUs, as well as other platforms. The changes are documented in the developer guide and signed in the release attestations.

## The facts

- Release date: 2026‑10‑17 ("Sat Oct 17 2026 23:11:34 GMT+0200")
- Adds Q1_0 and Q2_0 low‑bit formats for Hexagon NPU

## Why it matters

Developers can now run larger language models on Snapdragon‑based phones and tablets with lower latency and power use.

## Sources & references

1. [ggml-org/llama.cpp b11716](https://github.com/ggml-org/llama.cpp/releases/tag/b11716) – llama.cpp, 2026-10-17
2. [ggml-org/llama.cpp b11717](https://github.com/ggml-org/llama.cpp/releases/tag/b11717) – llama.cpp, 2026-10-18

Last updated: 2026-10-18