diff --git a/misc/llama-cpp/Makefile b/misc/llama-cpp/Makefile
index db42076ad4a8..102c3cf2498e 100644
--- a/misc/llama-cpp/Makefile
+++ b/misc/llama-cpp/Makefile
@@ -1,49 +1,50 @@
 PORTNAME=	llama-cpp
 DISTVERSIONPREFIX=	b
-DISTVERSION=	4120
+DISTVERSION=	4285
 CATEGORIES=	misc # machine-learning
 
 PATCH_SITES=	https://github.com/${GH_ACCOUNT}/${GH_PROJECT}/commit/
 PATCHFILES=	121f915a09c1117d34aff6e8faf6d252aaf11027.patch:-p1 # Add missing pthread includes: https://github.com/ggerganov/llama.cpp/pull/9258
+PATCHFILES+=	723342566862305c6707bf90b69f86625cf26620.patch:-p1 # prevent build freezing, https://github.com/ggerganov/llama.cpp/pull/10713
 
 MAINTAINER=	yuri@FreeBSD.org
 COMMENT=	Facebook's LLaMA model in C/C++ # '
 WWW=		https://github.com/ggerganov/llama.cpp
 
 LICENSE=	MIT
 LICENSE_FILE=	${WRKSRC}/LICENSE
 
 BROKEN_armv7=	clang crashes, see https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=278810
 BROKEN_i386=	compilation fails, see https://github.com/ggerganov/llama.cpp/issues/9545
 
 USES=		cmake:testing compiler:c++11-lang python:run shebangfix
 USE_LDCONFIG=	yes
 
 USE_GITHUB=	yes
 GH_ACCOUNT=	ggerganov
 GH_PROJECT=	llama.cpp
 GH_TUPLE=	nomic-ai:kompute:4565194:kompute/kompute
 
 SHEBANG_GLOB=	*.py
 
 CMAKE_ON=	BUILD_SHARED_LIBS
 CMAKE_OFF=	LLAMA_BUILD_TESTS
 CMAKE_TESTING_ON=	LLAMA_BUILD_TESTS
 
 OPTIONS_DEFINE=		EXAMPLES VULKAN
 OPTIONS_DEFAULT=	VULKAN
 OPTIONS_SUB=		yes
 
 EXAMPLES_CMAKE_BOOL=	LLAMA_BUILD_EXAMPLES
 
 VULKAN_DESC=		Vulkan GPU offload support
 VULKAN_CMAKE_BOOL=	GGML_VULKAN
 VULKAN_BUILD_DEPENDS=	glslc:graphics/shaderc \
 			vulkan-headers>0:graphics/vulkan-headers
 VULKAN_LIB_DEPENDS=	libvulkan.so:graphics/vulkan-loader
 
 BINARY_ALIAS=	git=false
 
 # 2 tests fail: https://github.com/ggerganov/llama.cpp/issues/8906
 
 .include <bsd.port.mk>
diff --git a/misc/llama-cpp/distinfo b/misc/llama-cpp/distinfo
index e8afef3b44f2..6dfb0c90f552 100644
--- a/misc/llama-cpp/distinfo
+++ b/misc/llama-cpp/distinfo
@@ -1,7 +1,9 @@
-TIMESTAMP = 1731907679
-SHA256 (ggerganov-llama.cpp-b4120_GH0.tar.gz) = ff1e6cde07e3f2a587978ea58d54bece296b61055b500898f702d8fbeff52e73
-SIZE (ggerganov-llama.cpp-b4120_GH0.tar.gz) = 19557501
+TIMESTAMP = 1733639498
+SHA256 (ggerganov-llama.cpp-b4285_GH0.tar.gz) = 5ddcac4db11002f50940b241b289e5bfca81c5032332427c622819dccf61e717
+SIZE (ggerganov-llama.cpp-b4285_GH0.tar.gz) = 19423062
 SHA256 (nomic-ai-kompute-4565194_GH0.tar.gz) = 95b52d2f0514c5201c7838348a9c3c9e60902ea3c6c9aa862193a212150b2bfc
 SIZE (nomic-ai-kompute-4565194_GH0.tar.gz) = 13540496
 SHA256 (121f915a09c1117d34aff6e8faf6d252aaf11027.patch) = 9a0c47ae3cb7dd51b6ce19187dafd48578210f69558f7c8044ee480471f1fd33
 SIZE (121f915a09c1117d34aff6e8faf6d252aaf11027.patch) = 591
+SHA256 (723342566862305c6707bf90b69f86625cf26620.patch) = 20074686fe70eb702528fdeb26b98b2ea81589529365cacf34edb21482040f70
+SIZE (723342566862305c6707bf90b69f86625cf26620.patch) = 7101
diff --git a/misc/llama-cpp/pkg-plist b/misc/llama-cpp/pkg-plist
index dd687531ab97..bf60b248b492 100644
--- a/misc/llama-cpp/pkg-plist
+++ b/misc/llama-cpp/pkg-plist
@@ -1,60 +1,63 @@
 bin/convert_hf_to_gguf.py
-bin/llama-batched
-bin/llama-batched-bench
-bin/llama-bench
-bin/llama-cli
-bin/llama-convert-llama2c-to-ggml
-bin/llama-cvector-generator
-bin/llama-embedding
-bin/llama-simple-chat
+%%EXAMPLES%%bin/llama-batched
+%%EXAMPLES%%bin/llama-batched-bench
+%%EXAMPLES%%bin/llama-bench
+%%EXAMPLES%%bin/llama-cli
+%%EXAMPLES%%bin/llama-convert-llama2c-to-ggml
+%%EXAMPLES%%bin/llama-cvector-generator
+%%EXAMPLES%%bin/llama-embedding
 %%EXAMPLES%%bin/llama-eval-callback
 %%EXAMPLES%%bin/llama-export-lora
 %%EXAMPLES%%bin/llama-gbnf-validator
 %%EXAMPLES%%bin/llama-gguf
 %%EXAMPLES%%bin/llama-gguf-hash
 %%EXAMPLES%%bin/llama-gguf-split
 %%EXAMPLES%%bin/llama-gritlm
 %%EXAMPLES%%bin/llama-imatrix
 %%EXAMPLES%%bin/llama-infill
 %%EXAMPLES%%bin/llama-llava-cli
 %%EXAMPLES%%bin/llama-lookahead
 %%EXAMPLES%%bin/llama-lookup
 %%EXAMPLES%%bin/llama-lookup-create
 %%EXAMPLES%%bin/llama-lookup-merge
 %%EXAMPLES%%bin/llama-lookup-stats
 %%EXAMPLES%%bin/llama-minicpmv-cli
 %%EXAMPLES%%bin/llama-parallel
 %%EXAMPLES%%bin/llama-passkey
 %%EXAMPLES%%bin/llama-perplexity
 %%EXAMPLES%%bin/llama-quantize
 %%EXAMPLES%%bin/llama-quantize-stats
 %%EXAMPLES%%bin/llama-retrieval
+%%EXAMPLES%%bin/llama-run
 %%EXAMPLES%%bin/llama-save-load-state
 %%EXAMPLES%%bin/llama-server
 %%EXAMPLES%%bin/llama-simple
+%%EXAMPLES%%bin/llama-simple-chat
 %%EXAMPLES%%bin/llama-speculative
+%%EXAMPLES%%bin/llama-speculative-simple
 %%EXAMPLES%%bin/llama-tokenize
 %%VULKAN%%bin/vulkan-shaders-gen
 include/ggml-alloc.h
 include/ggml-backend.h
 include/ggml-blas.h
 include/ggml-cann.h
 include/ggml-cpu.h
 include/ggml-cuda.h
 include/ggml-kompute.h
 include/ggml-metal.h
 include/ggml-opt.h
 include/ggml-rpc.h
 include/ggml-sycl.h
 include/ggml-vulkan.h
 include/ggml.h
+include/llama-cpp.h
 include/llama.h
 lib/cmake/llama/llama-config.cmake
 lib/cmake/llama/llama-version.cmake
-lib/libggml.so
 lib/libggml-base.so
 lib/libggml-cpu.so
 lib/libggml-vulkan.so
+lib/libggml.so
 lib/libllama.so
 %%EXAMPLES%%lib/libllava_shared.so
 libdata/pkgconfig/llama.pc