From e32d0f50deeb9b47830570e2d6cb21c02de72a7b Mon Sep 17 00:00:00 2001
From: localai-org-maint-bot
 <306269227+localai-org-maint-bot@users.noreply.github.com>
Date: Wed, 7 Oct 2026 04:04:27 +0000
Subject: [PATCH] chore(gallery): add POCKET-Darwin-180B GGUF

Add the published four-shard UD-Q4_K_XL build for llama.cpp with
revision-pinned downloads and verified SHA256 checksums. Document
installation, download size, text-only input, and sampling defaults.

Assisted-by: Codex:gpt-6
---
 docs/content/features/model-gallery.md | 19 ++++++++++
 gallery/index.yaml                     | 51 ++++++++++++++++++++++++++
 2 files changed, 70 insertions(+)

diff --git a/docs/content/features/model-gallery.md b/docs/content/features/model-gallery.md
index 3bdfc6d41..5f44bc98c 100644
--- a/docs/content/features/model-gallery.md
+++ b/docs/content/features/model-gallery.md
@@ -39,6 +39,25 @@ Both views use the same model selection and store the view, search, filter, and
 selection in the URL. Installing from Explore does not move you away from the
 catalog; the entry updates in place when the operation finishes.
 
+## POCKET-Darwin-180B
+
+[POCKET-Darwin-180B](https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF)
+is a text-only reasoning model based on Qwen3.8-Flash-Next. Install its
+UD-Q4_K_XL build with:
+
+```bash
+local-ai models install pocket-darwin-180b-q4
+```
+
+The entry downloads all four GGUF shards, about 111.3 GB in total, and uses
+`llama-cpp`. Allow additional memory for inference; download size is not a
+runtime memory estimate. This build has no vision encoder.
+
+The default context is 8192 tokens. Sampling uses temperature 1.0, top-p 0.95,
+and top-k 20. Allow at least 2048 output tokens for reasoning. The publisher
+provides one GGUF quantization, so this entry has no alternate variants.
+The weights use the [Qwen Community License 1.0](https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF/blob/71f3b4d3a68af20812af27b2e611501ab43baf1e/LICENSE).
+
 ## Cyber-Ornith 1.5 9B
 
 Cyber-Ornith 1.5 is a Qwen3.5 fine-tune for security auditing, terminal tasks, and tool use.
diff --git a/gallery/index.yaml b/gallery/index.yaml
index 043e1542c..cdb371932 100644
--- a/gallery/index.yaml
+++ b/gallery/index.yaml
@@ -1751,6 +1751,57 @@
     - filename: llama-cpp/models/llm-jp-4-33b-thinking/llm-jp-4-33b-thinking-BF16.gguf
       uri: huggingface://llm-jp/llm-jp-4-33b-thinking-gguf/llm-jp-4-33b-thinking-BF16.gguf
       sha256: 7bb8465702b5c4a5d94e03e62921d5917181a2eda56da31edfdcd47fab2e9964
+- name: pocket-darwin-180b-q4
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/FINAL-Bench/Darwin-180B-RSI-R3
+    - https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF
+  description: |
+    POCKET-Darwin-180B is a text-only reasoning model based on
+    Qwen3.8-Flash-Next. This UD-Q4_K_XL GGUF build uses llama.cpp and
+    downloads all four shards (about 111.3 GB). It uses the embedded
+    chat template, an 8192-token default context, and the publisher's
+    recommended sampling settings. Weights use Qwen Community License 1.0.
+  license: other
+  tags:
+    - llm
+    - gguf
+    - cpu
+    - gpu
+    - qwen
+    - moe
+    - reasoning
+  overrides:
+    backend: llama-cpp
+    context_size: 8192
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/pocket-darwin-180b/POCKET-Darwin-180B-UD-Q4_K_XL-00001-of-00004.gguf
+      temperature: 1
+      top_p: 0.95
+      top_k: 20
+      min_p: 0
+      repeat_penalty: 1
+      presence_penalty: 0
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/pocket-darwin-180b/POCKET-Darwin-180B-UD-Q4_K_XL-00001-of-00004.gguf
+      sha256: 8dd7789f02ff62e07f3a4dbd546d21fa6dc45cdbaac367631571d80cac8206a5
+      uri: https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF/resolve/71f3b4d3a68af20812af27b2e611501ab43baf1e/POCKET-Darwin-180B-UD-Q4_K_XL-00001-of-00004.gguf
+    - filename: llama-cpp/models/pocket-darwin-180b/POCKET-Darwin-180B-UD-Q4_K_XL-00002-of-00004.gguf
+      sha256: 8d2a48cace8791b2655173f8b507eb985471834511908006da6b4166b2155794
+      uri: https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF/resolve/71f3b4d3a68af20812af27b2e611501ab43baf1e/POCKET-Darwin-180B-UD-Q4_K_XL-00002-of-00004.gguf
+    - filename: llama-cpp/models/pocket-darwin-180b/POCKET-Darwin-180B-UD-Q4_K_XL-00003-of-00004.gguf
+      sha256: 0adc56c67b89cbfcaa67780e78fc0a15666ef2a1ebf5cdc51611fc1198ae256f
+      uri: https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF/resolve/71f3b4d3a68af20812af27b2e611501ab43baf1e/POCKET-Darwin-180B-UD-Q4_K_XL-00003-of-00004.gguf
+    - filename: llama-cpp/models/pocket-darwin-180b/POCKET-Darwin-180B-UD-Q4_K_XL-00004-of-00004.gguf
+      sha256: 0f45b1445b2a4d032d44766936107031d2f503b6bc7eb7fcf80f35232477ab5b
+      uri: https://huggingface.co/FINAL-Bench/POCKET-Darwin-180B-GGUF/resolve/71f3b4d3a68af20812af27b2e611501ab43baf1e/POCKET-Darwin-180B-UD-Q4_K_XL-00004-of-00004.gguf
+
 - name: huihui-qwen3.8-flash-next-abliterated-q4
   url: github:mudler/LocalAI/gallery/virtual.yaml@master
   urls: