Repository navigation

Release list

llama.cpp v0.7.0 with CUDA Latest

Latest

github-actions released this 16 Oct 03:36

v0.7.0

521f88e

llama.cpp v0.7.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.7.0
Commit: 620d2225cc436ab0b001e3200e590e3b286ef9ab

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.7.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.7.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.6.0 with CUDA

github-actions released this 06 Oct 05:12

v0.6.0

521f88e

llama.cpp v0.6.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.6.0
Commit: 8345f33

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.6.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.6.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.5.0 with CUDA

github-actions released this 24 Sep 03:16

v0.5.0

521f88e

llama.cpp v0.5.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.5.0
Commit: d2e5458

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.5.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.5.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.4.1 with CUDA

github-actions released this 15 Sep 03:28

v0.4.1

521f88e

llama.cpp v0.4.1 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.4.1
Commit: 391fac1

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.4.1-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.4.1-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.4.0 with CUDA

github-actions released this 05 Sep 03:34

v0.4.0

521f88e

llama.cpp v0.4.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.4.0
Commit: 427291b

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.4.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.4.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.3.0 with CUDA

github-actions released this 26 Aug 01:49

v0.3.0

521f88e

llama.cpp v0.3.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.3.0
Commit: c1d0e7a

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.3.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.3.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp v0.2.0 with CUDA

github-actions released this 22 Aug 01:54

v0.2.0

521f88e

llama.cpp v0.2.0 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/v0.2.0
Commit: 5a32f7b

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-v0.2.0-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-v0.2.0-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp b9993 with CUDA

github-actions released this 14 Jul 02:49

b9993

0ce85b5

llama.cpp b9993 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/b9993
Commit: 2969d6d

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-b9993-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-b9993-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp b9982 with CUDA

github-actions released this 13 Jul 03:04

b9982

0ce85b5

llama.cpp b9982 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/b9982
Commit: 99f3dc3

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-b9982-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-b9982-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

llama.cpp b9968 with CUDA

github-actions released this 12 Jul 03:08

b9968

0ce85b5

llama.cpp b9968 with CUDA Support

Pre-built binaries of llama.cpp with CUDA support for multiple CUDA versions.

Source: https://github.com/ggml-org/llama.cpp/releases/tag/b9968
Commit: 1d1d9a9

CUDA Versions

CUDA 12.8 - GPU compute capabilities: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0

Host architectures

Tarballs are published per host CPU architecture (Linux):

-amd64.tar.gz — x86_64 (most desktops, servers, cloud VMs)
-arm64.tar.gz — aarch64 (Grace Hopper / Grace Blackwell / DGX Spark / Ampere Altra)

GPU compute capability reference

7.5: Tesla T4, RTX 20xx series, Quadro RTX
8.0: A100
8.6: RTX 3000 series
8.9: RTX 4000 series, L4, L40
9.0: H100, H200, GH200
10.0: B200, GB200
12.0: RTX Pro series, RTX 50xx

Usage

Download the tarball matching your host CPU arch and CUDA version, then extract:

# amd64 host
tar -xzf llama.cpp-b9968-cuda-12.8-amd64.tar.gz
# arm64 host (e.g. Grace Blackwell)
tar -xzf llama.cpp-b9968-cuda-12.8-arm64.tar.gz
./llama-cli --help

Assets 4

Releases: JL2718/llama.cpp-cuda

Release list