Merge 7b1088cc80 into f58a864951

pyproject.toml

@@ -42,6 +42,7 @@ dependencies = [
   "pathvalidate",
   "charset-normalizer",
   "openai",
+  "ollama"
 ]
 
 [project.urls]

src/markitdown/_markitdown.py

@@ -13,6 +13,7 @@ import sys
 import tempfile
 import traceback
 import zipfile
+from http.client import responses
 from xml.dom import minidom
 from typing import Any, Dict, List, Optional, Union
 from pathlib import Path
@@ -1096,7 +1097,26 @@ class ImageConverter(MediaConverter):
                 content_type = "image/jpeg"
             image_base64 = base64.b64encode(image_file.read()).decode("utf-8")
             data_uri = f"data:{content_type};base64,{image_base64}"
+        # check if Ollama client
+        if str(type(client)) == "<class 'ollama._client.Client'>":
+            messages = [
+                {
+                    "role": "user",
+                    "content": prompt,
+                    'images': [local_path]
 
+                }
+            ]
+
+            response = client.chat(
+            model = model,
+            messages = messages,
+
+        )
+
+            return response.message.content
+
+        else:# use openai
             messages = [
                 {
                     "role": "user",
@@ -1111,7 +1131,6 @@ class ImageConverter(MediaConverter):
                     ],
                 }
             ]
-
             response = client.chat.completions.create(model=model, messages=messages)
             return response.choices[0].message.content