Models & Labs
Llama.cpp b10285 Release Enhances Deepseek-OCR
llama.cpp Releaseshigh confidence
Why it matters
- →Multi-row batching improves efficiency in OCR tasks.
- →Expanded platform support increases accessibility for developers.
- →Enhancements make llama.cpp more versatile for AI applications.
The b10285 release of llama.cpp brings notable updates, particularly for deepseek-ocr, with the addition of multi-row batching support. This change allows for more efficient processing by handling multiple rows simultaneously, rather than one at a time. The release also includes builds for various platforms, such as ROCm 7.2 on Ubuntu and CUDA 13 on Windows, broadening its compatibility. These updates enhance the tool's utility for developers working on OCR and AI applications, though no new model architectures are introduced.
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 b9258 Release Enhances Image Processing
llama.cpp Releases · May 21, 2026 · Same story
- 2llama.cpp b10005 Release Enhances Compatibility
llama.cpp Releases · July 15, 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.
