b10031

ggml-org/llama.cppb10031Jul 15, 2026by github-actions[bot]same day as b10025

What changed since b10025

low risk

b10031 only relaxes the tokenize CLI's --stdin mutual-exclusion check; the CUDA contiguity change was already in b10025.

New features

  • The tokenize tool now accepts --stdin alongside flags it previously rejected as mutually exclusive, matching shell completion behavior.

Improvements

  • CLI argument validation for tokenize is now consistent with what the completion scripts advertise.

Upgrade notes

This is a small behavior relaxation in the tokenize utility; no config or model changes are needed. Scripts or wrappers that depended on the CLI erroring out when --stdin was combined with other flags should be re-checked, since that validation no longer fires.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

Removes a mutual-exclusion check for the `--stdin` tokenization flag to align CLI behavior with completion scripts.

Key Highlights

  • Drops `--stdin` mutual-exclusion check to match CLI and completion behavior
  • Standard release binaries for all supported platforms

Release Notes (excerpt)

<details open> tokenize : drop --stdin mutual-exclusion check (#25672) match cli and completion, which don't enforce it </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) - Ubuntu x64 (SYCL FP16)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories