b10219

ggml-org/llama.cppb10219Aug 1, 2026by github-actions[bot]same day as b10218

What changed since b10218

low risk

b10219 is a single CLI bug fix: reasoning content is now persisted in chat history, not just displayed while streaming.

Improvements

  • llama-cli now stores streamed reasoning alongside assistant content in the message history, so --reasoning-preserve can re-inject prior thoughts on later turns.

Bug fixes

  • Fixed llama-cli dropping reasoning_content from chat history, which previously made reasoning preservation across turns ineffective.

Upgrade notes

This is a drop-in patch release with no API, config, or GGUF format changes; existing flags such as --reasoning-preserve now behave as intended. No migration is required, though anyone relying on reasoning carry-over in CLI sessions should upgrade to get the corrected behavior.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

Fixes a bug in the CLI where reasoning content was displayed during streaming but not persisted to the chat history.

Key Highlights

  • CLI chat history persistence fix
  • Reasoning content preservation across turns
  • llama-cli stream display correction

New Features

  • CLI functionality improvement

Release Notes (excerpt)

<details open> cli : persist reasoning_content in chat history (#26362) * cli : persist reasoning_content in chat history llama-cli collected reasoning from the stream for display but only stored assistant content in messages, so --reasoning-preserve could not re-inject prior thoughts on later turns. </details> **Website:** - <https://llama.app> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories