b9837

ggml-org/llama.cppb9837Jun 29, 2026by github-actions[bot]same day as b9835

What changed since b9835

low risk

b9837 adds an opt-in --reasoning-preserve flag for Jinja/chat output, with no other changes noted since b9835.

New features

  • New --reasoning-preserve CLI flag that keeps reasoning content in chat output generated through Jinja templates.

Improvements

  • Help text for the new reasoning-preserve option was corrected before release.

Upgrade notes

The flag is opt-in, so existing invocations behave exactly as before unless you add --reasoning-preserve. No config migration or data format change is required; simply rebuild or pull the new binaries if you want the option. The single-model stop/reasoning UI fix from b9835 remains in place.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

Adds a new flag to preserve reasoning content when generating chat outputs using Jinja templates.

Key Highlights

  • Added `--reasoning-preserve` flag to Jinja and chat modules
  • Allows users to retain reasoning content in the final output
  • Corrected help message for the new flag

Release Notes (excerpt)

<details open> jinja, chat: add --reasoning-preserve flag (#25105) * jinja, chat: add --reasoning-preserve flag * correct help message </details> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework **Linux:** - Ubuntu x64 (CPU) - Ubuntu arm64 (CPU) - Ubuntu s390x (CPU) - Ubuntu x64 (Vulkan) - Ubuntu arm64 (Vulkan) - Ubuntu x64 (ROCm 7.2) - Ubuntu x64 (OpenVINO) - Ubuntu x64 (SYCL FP32)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories