# llama.cpp b11777 release adds updated Qwen output format and new binaries

> The open‑source llama.cpp library released version b11777 on Oct 21, 2026, fixing Qwen output reshaping and publishing pre‑built binaries for macOS, iOS and Linux.

Oossa · 2026-10-21 · https://oossa.com/en/llama-cpp-b11777-release-adds-updated-qwen-output-format-and-new-binaries

The ggml‑org team published llama.cpp version b11777. The update reshapes Qwen model output from a 3‑dimensional to a 2‑dimensional tensor, removing an unnecessary final reshape step. The release also provides ready‑to‑run binaries for macOS (Apple Silicon and Intel), iOS, and several Ubuntu builds including Vulkan and CUDA options.

## The facts

- Release tag: b11777
- Published: 2026-10-21

## Why it matters

Developers can download the new binaries and run local LLMs on more devices without building from source.

## Sources & references

1. [ggml-org/llama.cpp b11777](https://github.com/ggml-org/llama.cpp/releases/tag/b11777) – llama.cpp, 2026-10-21

Last updated: 2026-10-21