[Package Request]: Ollama.Ollama.Portable

[jjj-dw]

How can we help?

I would like someone else to build the manifest.

Please read and ensure the following

• The installer meets the above requirements

Please provide the following information

Download Page Url:https://github.com/ollama/ollama/
Publisher:ollama
Package Name:Ollama Portable
Description:Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Package Version:0.20.0
Installer Url:https://github.com/ollama/ollama/releases/download/v0.20.0/ollama-windows-amd64.zip

[Dragon1573]

Hi @jjj-dw, 👋🏼

could you please explain why we need an extra portable version?

15:30:34 D:\...\winget-pkgs  [master ≡] 0ms pwsh
$ winget.exe search -s winget Ollama.Ollama
名称   ID            版本
---------------------------
Ollama Ollama.Ollama 0.20.0

15:30:39 D:\...\winget-pkgs  [master ≡] 235ms pwsh
$ winget.exe show -s winget Ollama.Ollama
已找到 Ollama [Ollama.Ollama]
版本: 0.20.0
发布者: Ollama
发布服务器 URL: https://ollama.com/
发布服务器支持 URL: https://github.com/ollama/ollama/issues
描述: 在本地启动并运行大型语言模型
主页: https://ollama.com/
许可证: MIT
许可证 URL: https://github.com/ollama/ollama/blob/HEAD/LICENSE
版权所有: Copyright (c) Ollama
发行说明:
  Gemma 4
  Effective 2B (E2B)
  ollama run gemma4:e2b
  Effective 4B (E4B)
  ollama run gemma4:e4b
  26B (Mixture of Experts model with 4B active parameters)
  ollama run gemma4:26b
  31B (Dense)
  ollama run gemma4:31b
  What's Changed
  - docs: update pi docs by @ParthSareen in https://github.com/ollama/ollama/pull/15152
  - mlx: respect tokenizer add_bos_token setting in pipeline by @dhiltgen in https://github.com/ollama/ollama/pull/15185
  - tokenizer: add SentencePiece-style BPE support by @dhiltgen in https://github.com/ollama/ollama/pull/15162      
  Full Changelog: https://github.com/ollama/ollama/compare/v0.19.0...v0.20.0-rc0
发行说明 URL: https://github.com/ollama/ollama/releases/tag/v0.20.0
文档:
  文档: https://docs.ollama.com/
标记：
  llama
  llm
  mistral
  人工智能
  大语言模型
  深度求索
  聊天机器人
  通义千问
安装：
  安装程序类型： inno
  安装程序 URL： https://github.com/ollama/ollama/releases/download/v0.20.0/OllamaSetup.exe
  安装程序 SHA256： 15f5a318e08ef94842adc2c76496cb47b5dbfa2efca99c90923cbe04c09a4899
  发布日期: 2026-04-02
  支持脱机分发: true

[jjj-dw]

Hi @jjj-dw, 👋🏼

could you please explain why we need an extra portable version?

15:30:34 D:\...\winget-pkgs  [master ≡] 0ms pwsh
$ winget.exe search -s winget Ollama.Ollama
名称   ID            版本
---------------------------
Ollama Ollama.Ollama 0.20.0

15:30:39 D:\...\winget-pkgs  [master ≡] 235ms pwsh
$ winget.exe show -s winget Ollama.Ollama
已找到 Ollama [Ollama.Ollama]
版本: 0.20.0
发布者: Ollama
发布服务器 URL: https://ollama.com/
发布服务器支持 URL: https://github.com/ollama/ollama/issues
描述: 在本地启动并运行大型语言模型
主页: https://ollama.com/
许可证: MIT
许可证 URL: https://github.com/ollama/ollama/blob/HEAD/LICENSE
版权所有: Copyright (c) Ollama
发行说明:
  Gemma 4
  Effective 2B (E2B)
  ollama run gemma4:e2b
  Effective 4B (E4B)
  ollama run gemma4:e4b
  26B (Mixture of Experts model with 4B active parameters)
  ollama run gemma4:26b
  31B (Dense)
  ollama run gemma4:31b
  What's Changed
  - docs: update pi docs by @ParthSareen in https://github.com/ollama/ollama/pull/15152
  - mlx: respect tokenizer add_bos_token setting in pipeline by @dhiltgen in https://github.com/ollama/ollama/pull/15185
  - tokenizer: add SentencePiece-style BPE support by @dhiltgen in https://github.com/ollama/ollama/pull/15162      
  Full Changelog: https://github.com/ollama/ollama/compare/v0.19.0...v0.20.0-rc0
发行说明 URL: https://github.com/ollama/ollama/releases/tag/v0.20.0
文档:
  文档: https://docs.ollama.com/
标记：
  llama
  llm
  mistral
  人工智能
  大语言模型
  深度求索
  聊天机器人
  通义千问
安装：
  安装程序类型： inno
  安装程序 URL： https://github.com/ollama/ollama/releases/download/v0.20.0/OllamaSetup.exe
  安装程序 SHA256： 15f5a318e08ef94842adc2c76496cb47b5dbfa2efca99c90923cbe04c09a4899
  发布日期: 2026-04-02
  支持脱机分发: true

because it's only contain Ollama CLI，no GUI，use less ram or rom

[Dragon1573]

17:25:56 D:\...\winget-pkgs  [master ≡] 1ms pwsh
$ gh release view --repo ollama/ollama --json assets --jq '.assets[]' |
> ConvertFrom-Json |
> Select-Object -Property url, digest |
> Where-Object { $_.url -cmatch 'windows' -or $_.url -cmatch '\.exe$' } |
> Format-List

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64-mlx.zip
digest : sha256:c3a1255a22d9965710c63120fd5cce960a7cf954f8d691195f1aa3074dd36fce

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64-rocm.zip
digest : sha256:41e9811f8cd2473cdfe416b4aa2ac5ceeb354df63a94fd0edd37309e39459c4f

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64.zip
digest : sha256:f26ead8baa55e919828ffe88cddf3e588444891e43e19a059f1d6c933fb1a5b6

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-arm64.zip
digest : sha256:e4d2111aa2b1bed7768b0dd20c17bc7d2860f0ee7b8c78bffca5920504419291

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/OllamaSetup.exe
digest : sha256:31358c5062d608353a69f0f0ef3006dc397b7017819b9cb412a9a96ed2aa0bcc

There're 2 other variants of Windows prebuilt binaries. Should we also add them as Ollama.Ollama.Portable.MLX and Ollama.Ollama.Portable.ROCm?

[jjj-dw]

17:25:56 D:\...\winget-pkgs  [master ≡] 1ms pwsh
$ gh release view --repo ollama/ollama --json assets --jq '.assets[]' |
> ConvertFrom-Json |
> Select-Object -Property url, digest |
> Where-Object { $_.url -cmatch 'windows' -or $_.url -cmatch '\.exe$' } |
> Format-List

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64-mlx.zip
digest : sha256:c3a1255a22d9965710c63120fd5cce960a7cf954f8d691195f1aa3074dd36fce

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64-rocm.zip
digest : sha256:41e9811f8cd2473cdfe416b4aa2ac5ceeb354df63a94fd0edd37309e39459c4f

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-amd64.zip
digest : sha256:f26ead8baa55e919828ffe88cddf3e588444891e43e19a059f1d6c933fb1a5b6

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/ollama-windows-arm64.zip
digest : sha256:e4d2111aa2b1bed7768b0dd20c17bc7d2860f0ee7b8c78bffca5920504419291

url    : https://github.com/ollama/ollama/releases/download/v0.20.2/OllamaSetup.exe
digest : sha256:31358c5062d608353a69f0f0ef3006dc397b7017819b9cb412a9a96ed2aa0bcc

There're 2 other variants of Windows prebuilt binaries. Should we also add them as and ?Ollama.Ollama.Portable.MLX``Ollama.Ollama.Portable.ROCm

AMD GPU: ollama-windows-amd64-rocm.zip
MLX (CUDA): ollama-windows-amd64-mlx.zip

[Dragon1573]

Warning

The ollama-windows-amd64-rocm.zip and ollama-windows-amd64-mlx.zip should not be introduced as another package!

According to the compressed archived README_ROCm.txt, it's an external library "addon" for ollama-windows-amd64.zip main executable bundle.