# llama.cpp b12019 adjusts Qwen3-Coder reasoning output

> The llama.cpp release changes how Qwen3-Coder think blocks are handled when reasoning_format is set to none, and offers downloads for several platforms.

Oossa · 2026-10-31 · https://oossa.com/en/llama-cpp-b12019-adjusts-qwen3-coder-reasoning-output

The ggml-org team has released llama.cpp b12019. It updates chat handling for Qwen3-Coder: with reasoning_format set to none, the software can accept two think-block variants and strip the block from the response.

The release also includes ready-to-download builds for macOS, iOS and several Ubuntu systems.

## The facts

- Release: llama.cpp b12019.
- The change covers two Qwen3-Coder think-block variants when reasoning_format is none.

## Why it matters

People running Qwen3-Coder through llama.cpp can use the updated handling to avoid think blocks appearing in responses configured to omit reasoning, though the release notes do not specify which applications or setups are affected.

## Sources & references

1. [ggml-org/llama.cpp b12019](https://github.com/ggml-org/llama.cpp/releases/tag/b12019) – llama.cpp, 2026-10-31

Last updated: 2026-10-31