# llama.cpp release b11701 adds Beam (Reflection) feature

> The open‑source Llama.cpp library released version b11701 on Oct 17, 2026, adding a Beam (Reflection) model option and new binaries for macOS, iOS, Linux, Vulkan and CUDA.

Oossa · 2026-10-16 · https://oossa.com/en/llama-cpp-release-b11701-adds-beam-reflection-feature

The ggml‑org team published llama.cpp version b11701 on Oct 17, 2026. The update adds a Beam (Reflection) model feature, which lets users run beam‑search style inference. It also provides pre‑built binaries for macOS Apple Silicon, Intel, iOS, and various Ubuntu builds including Vulkan and CUDA 12.8 support.

## The facts

- Release version b11701 announced on 2026‑10‑17
- Adds Beam (Reflection) model option

## Why it matters

Developers can now use llama.cpp with beam search on more platforms, improving the quality of generated text without writing custom code.

## Sources & references

1. [ggml-org/llama.cpp b11701](https://github.com/ggml-org/llama.cpp/releases/tag/b11701) – llama.cpp, 2026-10-16

Last updated: 2026-10-16