diff --git a/.gitignore b/.gitignore
index c6b115c56..b1d549014 100644
--- a/.gitignore
+++ b/.gitignore
@@ -149,4 +149,6 @@ dump.rdb
 .DS_Store
 
 # Tester app for unit tests
-scripts/vespa_local/vespa_tester_app.zip
\ No newline at end of file
+scripts/vespa_local/vespa_tester_app.zip
+
+src/marqo/tensor_search/cache_dir/*
\ No newline at end of file
diff --git a/RELEASE.md b/RELEASE.md
index 1fb75a5f2..3549ae920 100644
--- a/RELEASE.md
+++ b/RELEASE.md
@@ -1,3 +1,16 @@
+# Release 2.10.2
+
+ ## Bug fixes and minor changes
+ - Fix an issue where CUDA was not automatically selected as the default device for the `embed` endpoint, even when available [#941](https://github.com/marqo-ai/marqo/pull/941).
+
+# Release 2.10.1
+
+## Bug fixes and minor changes
+- Improve the clarity of the error message when Marqo can not download the provided image ([#905](https://github.com/marqo-ai/marqo/pull/905)).
+- Improve the error message in hybrid search to avoid confusion ([#900](https://github.com/marqo-ai/marqo/pull/900)).
+- Fix a bug where a `500` error is returned when an unsupported search method is provided. Marqo now correctly returns a `400` error ([#899](https://github.com/marqo-ai/marqo/pull/899)).
+- Fix a bug where a `500` error is returned when an invalid image URL with non-ASCII characters is provided. Marqo now encodes the image URL correctly ([#908](https://github.com/marqo-ai/marqo/pull/908)).
+
 # Release 2.10.0
 
 ## New features
diff --git a/src/marqo/core/embed/embed.py b/src/marqo/core/embed/embed.py
index d8c00815c..fa8430a49 100644
--- a/src/marqo/core/embed/embed.py
+++ b/src/marqo/core/embed/embed.py
@@ -13,6 +13,7 @@
 from marqo.tensor_search.tensor_search_logging import get_logger
 from marqo.core.utils.prefix import determine_text_prefix, DeterminePrefixContentType
 from marqo.vespa.vespa_client import VespaClient
+from marqo.tensor_search import utils
 
 logger = get_logger(__name__)
 
@@ -64,7 +65,7 @@ def embed_content(
 
         # Set default device if not provided
         if device is None:
-            device = self.default_device
+            device = utils.read_env_vars_and_defaults("MARQO_BEST_AVAILABLE_DEVICE")
 
 
         # Content validation is done in API model layer
diff --git a/src/marqo/version.py b/src/marqo/version.py
index 23ca83757..0cbbb5e0c 100644
--- a/src/marqo/version.py
+++ b/src/marqo/version.py
@@ -1,4 +1,4 @@
-__version__ = "2.10.1"
+__version__ = "2.10.2"
 
 def get_version() -> str:
     return f"{__version__}"
diff --git a/tests/tensor_search/integ_tests/test_embed.py b/tests/tensor_search/integ_tests/test_embed.py
index 38e4e3527..eb07fd068 100644
--- a/tests/tensor_search/integ_tests/test_embed.py
+++ b/tests/tensor_search/integ_tests/test_embed.py
@@ -20,6 +20,7 @@
 from marqo.vespa.models.query_result import Root, Child, RootFields
 from marqo.tensor_search.models.private_models import S3Auth, ModelAuth, HfAuth
 from marqo.api.models.embed_request import EmbedRequest
+from marqo.tensor_search import utils
 import os
 import pprint
 import unittest
@@ -150,6 +151,45 @@ def tearDown(self) -> None:
         super().tearDown()
         self.device_patcher.stop()
 
+    def test_embed_content_cuda_device_as_default(self):
+         """
+         Test that embed_content uses the default device when no device is specified.
+         """
+         for index in [self.unstructured_default_text_index, self.structured_default_text_index]:
+             with self.subTest(index=index.type):
+                 expected_devices = ["cuda", "cpu"]
+                 for expected_device in expected_devices:
+                     with patch.dict(os.environ, {"MARQO_BEST_AVAILABLE_DEVICE": expected_device}):
+                         with patch('marqo.tensor_search.tensor_search.run_vectorise_pipeline') as mock_vectorise:
+                             mock_vectorise.return_value = {0: [0.1, 0.2, 0.3]}
+
+                             embed_res = embed(
+                                 marqo_config=self.config,
+                                 index_name=index.name,
+                                 embedding_request=EmbedRequest(
+                                     content=["This is a test document"]
+                                 ),
+                                 device=None
+                             )
+
+                         # Check that run_vectorise_pipeline was called
+                         mock_vectorise.assert_called_once()
+
+                         # Get the arguments passed to run_vectorise_pipeline
+                         args, kwargs = mock_vectorise.call_args
+
+                         # Print the args and kwargs for debugging
+                         print(f"args passed to run_vectorise_pipeline: {args}")
+                         print(f"kwargs passed to run_vectorise_pipeline: {kwargs}")
+
+                         # Check that the device passed to run_vectorise_pipeline matches the expected value
+                         self.assertEqual(args[2], expected_device)
+
+                         # Check the result
+                         self.assertEqual(embed_res["content"], ["This is a test document"])
+                         self.assertIsInstance(embed_res["embeddings"][0], list)
+                         self.assertEqual(embed_res["embeddings"][0], [0.1, 0.2, 0.3])
+
     def test_embed_equivalent_to_add_docs(self):
         """
         Ensure that the embedding returned by embed endpoint matches the one created by add_docs.