Models & Labs
Llama.cpp b10816 Release Enhances M3 Tuning
llama.cpp Releaseshigh confidence
Why it matters
- →Enhances performance for Apple Silicon users with new M3 tuning.
- →Maintains broad compatibility across multiple platforms.
- →Demonstrates ongoing optimization efforts for diverse hardware.
Llama.cpp has released version b10816, which includes new tuning parameters for Apple's M3 chips in the fa_vec_tuned_table. This update enhances performance on macOS devices by adding q4_0, q4_1, q5_0, and q5_1 tunings in the ggml-metal-tuning. The release continues to support a wide range of platforms, including Linux, Windows, and Android, ensuring broad compatibility. This update highlights llama.cpp's ongoing efforts to optimize its software for various hardware configurations.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1llama.cpp b10743 Release Enhances M2 Pro Tuning
llama.cpp Releases · September 2, 2026 · Same story
- 2llama.cpp b10739 Release Enhances M2 Max Performance
llama.cpp Releases · September 2, 2026 · Same story
More from llama.cpp Releases
Coding Toolscoding
llama.cpp b12022 adds ROCm 10 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally bringing ROCm 10.0 to Linux and Windows alongside CUDA 13.4, effectively closing the hardware gap for AMD users who were previously left behind. The inclusion of native Snapdragon support on Linux marks a significant step toward running local LLMs on ARM-based mobile chips without emulation overhead. While KleidiAI on Apple Silicon is disabled in this build, the broader expansion to diverse accelerators ensures that developers aren't locked into NVIDIA ecosystems anymore.
llama.cpp Releases
Coding Toolscoding
llama.cpp b12023 fixes shared prefix bug
This release addresses a specific inference artifact where decision prefixes were incorrectly bleeding into output embeddings. By capping shared prefixes at the first output position, llama.cpp ensures that option tokens remain distinct for parent and child tasks. This is a targeted fix for multi-step reasoning or voting workflows rather than a performance boost. It stabilizes complex generation patterns without changing the underlying model architecture.
