From 3e353c980b7c7eb7f5d7fde7baa0aaf91bc74208 Mon Sep 17 00:00:00 2001
From: cvrunmin <cvrunmin@outlook.com>
Date: Fri, 28 Nov 2025 14:14:20 +0800
Subject: [PATCH] docs: add infos for split gguf

---
 docs/api.md        |  3 ++-
 docs/import.mdx    |  4 ++++
 docs/modelfile.mdx | 12 ++++++++++--
 3 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/docs/api.md b/docs/api.md
index 99ceaa11a..9bab5dd46 100644
--- a/docs/api.md
+++ b/docs/api.md
@@ -1176,7 +1176,7 @@ Create a model from:
 
 - another model;
 - a safetensors directory; or
-- a GGUF file.
+- a GGUF file or directory.
 
 If you are creating a model from a safetensors directory or from a GGUF file, you must [create a blob](#create-a-blob) for each of the files and then use the file name and SHA256 digest associated with each blob in the `files` field.
 
@@ -1270,6 +1270,7 @@ A stream of JSON objects is returned:
 #### Create a model from GGUF
 
 Create a model from a GGUF file. The `files` parameter should be filled out with the file name and SHA256 digest of the GGUF file you wish to use. Use [/api/blobs/:digest](#push-a-blob) to push the GGUF file to the server before calling this API.
+For a model stored in multiple split GGUF files, includes all split GGUF files in the `files` parameter with the file names and SHA256 digests. It is recommended to provide files in split number order even though Ollama itself will sort them in order.
 
 ##### Request
 
diff --git a/docs/import.mdx b/docs/import.mdx
index b19596894..26a636564 100644
--- a/docs/import.mdx
+++ b/docs/import.mdx
@@ -88,6 +88,10 @@ To import a GGUF model, create a `Modelfile` containing:
 ```dockerfile
 FROM /path/to/file.gguf
 ```
+Or:
+```dockerfile
+FROM /path/to/gguf/split/directory
+```
 
 For a GGUF adapter, create the `Modelfile` with:
 
diff --git a/docs/modelfile.mdx b/docs/modelfile.mdx
index c91d7310c..1bc28c555 100644
--- a/docs/modelfile.mdx
+++ b/docs/modelfile.mdx
@@ -12,7 +12,7 @@ A Modelfile is the blueprint to create and share customized models using Ollama.
   - [FROM (Required)](#from-required)
     - [Build from existing model](#build-from-existing-model)
     - [Build from a Safetensors model](#build-from-a-safetensors-model)
-    - [Build from a GGUF file](#build-from-a-gguf-file)
+    - [Build from a GGUF file](#build-from-a-gguf-model)
   - [PARAMETER](#parameter)
     - [Valid Parameters and Values](#valid-parameters-and-values)
   - [TEMPLATE](#template)
@@ -129,7 +129,7 @@ Currently supported model architectures:
 - Gemma (including Gemma 1 and Gemma 2)
 - Phi3
 
-#### Build from a GGUF file
+#### Build from a GGUF model
 
 ```
 FROM ./ollama-model.gguf
@@ -137,6 +137,14 @@ FROM ./ollama-model.gguf
 
 The GGUF file location should be specified as an absolute path or relative to the `Modelfile` location.
 
+For GGUF model split into multiple files:
+
+```
+FROM <model directory>
+```
+
+The model directory should contain solely the split GGUF weights of one model.
+
 ### PARAMETER
 
 The `PARAMETER` instruction defines a parameter that can be set when the model is run.