From 9bc8bb07cf16c8a9116b8029e63d1015d150cccb Mon Sep 17 00:00:00 2001
From: Guillaume Calmettes <gcalmettes@scaleway.com>
Date: Thu, 5 Jun 2025 14:59:28 +0200
Subject: [PATCH] [Bugfix] properly catch PIL-related errors for vision models
 when incorrect data urls are provided (#19202)

Signed-off-by: Guillaume Calmettes <gcalmettes@scaleway.com>
---
 tests/multimodal/test_utils.py | 13 +++++++++++++
 vllm/multimodal/utils.py       | 30 +++++++++++++++++++-----------
 2 files changed, 32 insertions(+), 11 deletions(-)

diff --git a/tests/multimodal/test_utils.py b/tests/multimodal/test_utils.py
index e4debb47ce..c8a5448221 100644
--- a/tests/multimodal/test_utils.py
+++ b/tests/multimodal/test_utils.py
@@ -141,6 +141,19 @@ async def test_fetch_image_local_files(image_url: str):
                 f"file://{temp_dir}/../{os.path.basename(image_url)}")
 
 
+@pytest.mark.asyncio
+async def test_fetch_image_error_conversion():
+    connector = MediaConnector()
+    broken_img = "data:image/png;base64,aGVsbG9fdmxsbV9jb21tdW5pdHkK"
+
+    # PIL.UnidentifiedImageError should be converted to ValueError
+    with pytest.raises(ValueError):
+        await connector.fetch_image_async(broken_img)
+
+    with pytest.raises(ValueError):
+        connector.fetch_image(broken_img)
+
+
 @pytest.mark.asyncio
 @pytest.mark.parametrize("video_url", TEST_VIDEO_URLS)
 @pytest.mark.parametrize("num_frames", [-1, 32, 1800])
diff --git a/vllm/multimodal/utils.py b/vllm/multimodal/utils.py
index 2b34cdf40b..11a25f8515 100644
--- a/vllm/multimodal/utils.py
+++ b/vllm/multimodal/utils.py
@@ -9,7 +9,7 @@ from urllib.parse import ParseResult, urlparse
 import numpy as np
 import numpy.typing as npt
 import torch
-from PIL import Image
+from PIL import Image, UnidentifiedImageError
 
 import vllm.envs as envs
 from vllm.connections import HTTPConnection, global_http_connection
@@ -185,11 +185,15 @@ class MediaConnector:
         """
         image_io = ImageMediaIO(image_mode=image_mode)
 
-        return self.load_from_url(
-            image_url,
-            image_io,
-            fetch_timeout=envs.VLLM_IMAGE_FETCH_TIMEOUT,
-        )
+        try:
+            return self.load_from_url(
+                image_url,
+                image_io,
+                fetch_timeout=envs.VLLM_IMAGE_FETCH_TIMEOUT,
+            )
+        except UnidentifiedImageError as e:
+            # convert to ValueError to be properly caught upstream
+            raise ValueError(str(e)) from e
 
     async def fetch_image_async(
         self,
@@ -204,11 +208,15 @@ class MediaConnector:
         """
         image_io = ImageMediaIO(image_mode=image_mode)
 
-        return await self.load_from_url_async(
-            image_url,
-            image_io,
-            fetch_timeout=envs.VLLM_IMAGE_FETCH_TIMEOUT,
-        )
+        try:
+            return await self.load_from_url_async(
+                image_url,
+                image_io,
+                fetch_timeout=envs.VLLM_IMAGE_FETCH_TIMEOUT,
+            )
+        except UnidentifiedImageError as e:
+            # convert to ValueError to be properly caught upstream
+            raise ValueError(str(e)) from e
 
     def fetch_video(
         self,