Models & Labs
llama.cpp b11081 release with CUDA 13 and ROCm 10 support
llama.cpp Releaseshigh confidence
Why it matters
- →CUDA 13.4 support enables immediate compatibility with the latest NVIDIA GPU drivers and hardware.
- →ROCm 10.0 integration provides native performance optimization for modern AMD accelerators.
- →Pre-built binaries reduce friction for developers who cannot or prefer not to compile from source.
llama.cpp has released version b11081, expanding its pre-built binary support to include CUDA 13.4 and ROCm 10.0 for both Linux and Windows platforms. The update also maintains compatibility with existing CUDA 12 builds and continues to feature KleidiAI optimization for Apple Silicon Macs. Internal testing infrastructure has been refined to better handle tensor data standard deviation and recurrent state rollback checks, ensuring stability across diverse model architectures. This release allows developers to run local inference on the latest NVIDIA and AMD GPUs without compiling from source.
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.
What happened next
- 1llama.cpp b11222 release with ROCm 10 and CUDA 13
llama.cpp Releases · September 28, 2026 · Same story
- 2llama.cpp b12182 release with CUDA 13 and ROCm 10
llama.cpp Releases · November 7, 2026 · Same story
Follow this story
Open the full story →More from llama.cpp Releases
Coding Toolscoding
llama.cpp b12336 fixes HunyuanVL crash
This release quietly patches a critical memory overflow in the mtmd module that caused crashes when loading HunyuanVL models. The issue stemmed from F64 position embeddings overflowing an F32 buffer, a subtle bug that broke multimodal inference for users relying on this specific architecture. By enforcing strict F32 requirements, llama.cpp stabilizes support for Tencent's vision-language model without adding new features or changing the core API. It’s a necessary maintenance update that keeps local multimodal workflows from breaking unexpectedly.
llama.cpp Releases
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
