Download Latest Version textgen-portable-3.21-macos-arm64.zip (198.5 MB)
Email in envelope

Get an email when there's a new version of Text Generation Web UI

Home / v3.18
Name Modified Size InfoDownloads / Week
Parent folder
textgen-portable-3.18-windows-cuda12.4.zip 2025-11-19 909.2 MB
0
textgen-portable-3.18-windows-vulkan.zip 2025-11-19 213.4 MB
0
textgen-portable-3.18-windows-cpu.zip 2025-11-19 199.2 MB
0
textgen-portable-3.18-linux-cuda12.4.zip 2025-11-19 1.6 GB
0
textgen-portable-3.18-linux-rocm.zip 2025-11-19 614.9 MB
0
textgen-portable-3.18-linux-vulkan.zip 2025-11-19 292.2 MB
0
textgen-portable-3.18-linux-cpu.zip 2025-11-19 248.7 MB
0
textgen-portable-3.18-macos-arm64.zip 2025-11-19 192.1 MB
0
README.md 2025-11-19 1.2 kB
0
v3.18 source code.tar.gz 2025-11-19 24.9 MB
0
v3.18 source code.zip 2025-11-19 25.0 MB
0
Totals: 11 Items   4.3 GB 0

Changes

  • Add --cpu-moe flag for llama.cpp to move MoE model experts to CPU, reducing VRAM usage.
  • Add ROCm portable builds for AMD GPUs on Linux. This was made possible by PR https://github.com/oobabooga/llama-cpp-binaries/pull/7 by @ShortTimeNoSee. Thanks, @ShortTimeNoSee.
  • Remove deprecated macOS 13 wheels (no longer supported by GitHub Actions).

Backend updates

Portable builds

Below you can find self-contained packages that work with GGUF models (llama.cpp) and require no installation! Just download the right version for your system, unzip, and run.

Which version to download:

  • Windows/Linux:
  • NVIDIA GPU: Use cuda12.4.
  • AMD/Intel GPU: Use vulkan builds.

  • CPU only: Use cpu builds.

  • Mac:

  • Apple Silicon: Use macos-arm64.

Updating a portable install:

  1. Download and unzip the latest version.
  2. Replace the user_data folder with the one in your existing install. All your settings and models will be moved.
Source: README.md, updated 2025-11-19