diff --git a/libs/partners/openai/langchain_openai/chat_models/base.py b/libs/partners/openai/langchain_openai/chat_models/base.py
index 9e6d24c9b6e..9311ac4fff8 100644
--- a/libs/partners/openai/langchain_openai/chat_models/base.py
+++ b/libs/partners/openai/langchain_openai/chat_models/base.py
@@ -32,6 +32,7 @@ from urllib.parse import urlparse
 import certifi
 import openai
 import tiktoken
+from langchain_core._api.beta_decorator import warn_beta
 from langchain_core._api.deprecation import deprecated
 from langchain_core.callbacks import (
     AsyncCallbackManagerForLLMRun,
@@ -416,6 +417,7 @@ def _handle_openai_bad_request(e: openai.BadRequestError) -> None:
 
 
 _MessageContent = Union[str, list[Union[str, dict]]]
+WARNED_IMAGE_GEN_BETA = False
 
 
 def _get_image_bytes_from_content(
@@ -1279,10 +1281,20 @@ class BaseChatOpenAI(BaseChatModel):
         run_manager: Optional[CallbackManagerForLLMRun] = None,
         **kwargs: Any,
     ) -> ChatResult:
+        global WARNED_IMAGE_GEN_BETA
+        if not WARNED_IMAGE_GEN_BETA:
+            warn_beta(message="Image generation via ChatOpenAI is in beta.")
+            WARNED_IMAGE_GEN_BETA = True
+
         prompt = messages[-1].text()
-        images = []
-        for message in messages:
-            images.extend(_get_image_bytes_from_content(message.content))
+
+        # Get last set of images
+        for message in reversed(messages):
+            images = _get_image_bytes_from_content(message.content)
+            if images:
+                break
+        else:
+            images = []
         if images:
             result: ImagesResponse = self.root_client.images.edit(
                 model=self.model_name, image=images, prompt=prompt, **kwargs
@@ -1291,7 +1303,27 @@ class BaseChatOpenAI(BaseChatModel):
             result = self.root_client.images.generate(
                 model=self.model_name, prompt=prompt, **kwargs
             )
-        output_message = AIMessage(content="", response_metadata=result.model_dump())
+        image_blocks = []
+        if result.data:
+            for image in result.data:
+                if image.b64_json:
+                    image_blocks.append(
+                        {
+                            "type": "image",
+                            "source_type": "base64",
+                            "data": image.b64_json,
+                            "mime_type": "image/png",
+                        }
+                    )
+        if result.usage:
+            usage_metadata = _create_usage_metadata_responses(result.usage.model_dump())
+        else:
+            usage_metadata = None
+        output_message = AIMessage(
+            content=image_blocks or "",  # type: ignore[arg-type]
+            response_metadata={"created": result.created},
+            usage_metadata=usage_metadata,
+        )
         return ChatResult(generations=[ChatGeneration(message=output_message)])
 
     def _get_encoding_model(self) -> tuple[str, tiktoken.Encoding]:
diff --git a/libs/partners/openai/pyproject.toml b/libs/partners/openai/pyproject.toml
index c2468fd3c07..7afb66316c3 100644
--- a/libs/partners/openai/pyproject.toml
+++ b/libs/partners/openai/pyproject.toml
@@ -8,7 +8,7 @@ license = { text = "MIT" }
 requires-python = "<4.0,>=3.9"
 dependencies = [
     "langchain-core<1.0.0,>=0.3.53",
-    "openai<2.0.0,>=1.68.2",
+    "openai<2.0.0,>=1.76.0",
     "tiktoken<1,>=0.7",
 ]
 name = "langchain-openai"
diff --git a/libs/partners/openai/uv.lock b/libs/partners/openai/uv.lock
index f0e68ea97eb..a1f03a8f75f 100644
--- a/libs/partners/openai/uv.lock
+++ b/libs/partners/openai/uv.lock
@@ -463,7 +463,7 @@ wheels = [
 
 [[package]]
 name = "langchain-core"
-version = "0.3.53"
+version = "0.3.55"
 source = { editable = "../../core" }
 dependencies = [
     { name = "jsonpatch" },
@@ -571,7 +571,7 @@ typing = [
 [package.metadata]
 requires-dist = [
     { name = "langchain-core", editable = "../../core" },
-    { name = "openai", specifier = ">=1.68.2,<2.0.0" },
+    { name = "openai", specifier = ">=1.76.0,<2.0.0" },
     { name = "tiktoken", specifier = ">=0.7,<1" },
 ]
 
@@ -829,7 +829,7 @@ wheels = [
 
 [[package]]
 name = "openai"
-version = "1.68.2"
+version = "1.76.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "anyio" },
@@ -841,9 +841,9 @@ dependencies = [
     { name = "tqdm" },
     { name = "typing-extensions" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/3f/6b/6b002d5d38794645437ae3ddb42083059d556558493408d39a0fcea608bc/openai-1.68.2.tar.gz", hash = "sha256:b720f0a95a1dbe1429c0d9bb62096a0d98057bcda82516f6e8af10284bdd5b19", size = 413429 }
+sdist = { url = "https://files.pythonhosted.org/packages/84/51/817969ec969b73d8ddad085670ecd8a45ef1af1811d8c3b8a177ca4d1309/openai-1.76.0.tar.gz", hash = "sha256:fd2bfaf4608f48102d6b74f9e11c5ecaa058b60dad9c36e409c12477dfd91fb2", size = 434660 }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/fd/34/cebce15f64eb4a3d609a83ac3568d43005cc9a1cba9d7fde5590fd415423/openai-1.68.2-py3-none-any.whl", hash = "sha256:24484cb5c9a33b58576fdc5acf0e5f92603024a4e39d0b99793dfa1eb14c2b36", size = 606073 },
+    { url = "https://files.pythonhosted.org/packages/59/aa/84e02ab500ca871eb8f62784426963a1c7c17a72fea3c7f268af4bbaafa5/openai-1.76.0-py3-none-any.whl", hash = "sha256:a712b50e78cf78e6d7b2a8f69c4978243517c2c36999756673e07a14ce37dc0a", size = 661201 },
 ]
 
 [[package]]