diff --git a/.github/workflows/test_openai.yml b/.github/workflows/test_openai.yml
index 975d17b3fe..1fce9c133e 100644
--- a/.github/workflows/test_openai.yml
+++ b/.github/workflows/test_openai.yml
@@ -29,49 +29,49 @@ jobs:
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_returns_valid_first_message
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_returns_valid_first_message
 
     - name: Test model sends message with keyword
       id: test_keyword_message
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_returns_keyword
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_returns_keyword
 
     - name: Test model uses external tool correctly
       id: test_external_tool
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_uses_external_tool
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_uses_external_tool
 
     - name: Test model recalls chat memory
       id: test_chat_memory
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_recall_chat_memory
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_recall_chat_memory
 
     - name: Test model uses 'archival_memory_search' to find secret
       id: test_archival_memory_search
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_archival_memory_retrieval
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_archival_memory_retrieval
 
     - name: Test model uses 'archival_memory_insert' to insert archival memories
       id: test_archival_memory_insert
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_archival_memory_insert
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_archival_memory_insert
 
     - name: Test model can edit core memories
       id: test_core_memory
       env:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       run: |
-        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4_edit_core_memory
+        poetry run pytest -s -vv tests/test_endpoints.py::test_openai_gpt_4o_edit_core_memory
 
     - name: Test embedding endpoint
       id: test_embedding_endpoint
diff --git a/letta/llm_api/openai.py b/letta/llm_api/openai.py
index 231cde29b4..8277239a7e 100644
--- a/letta/llm_api/openai.py
+++ b/letta/llm_api/openai.py
@@ -126,6 +126,7 @@ def build_openai_chat_completions_request(
     openai_message_list = [
         cast_message_to_subtype(m.to_openai_dict(put_inner_thoughts_in_kwargs=llm_config.put_inner_thoughts_in_kwargs)) for m in messages
     ]
+
     if llm_config.model:
         model = llm_config.model
     else:
diff --git a/letta/memory.py b/letta/memory.py
index 474d18c52d..a873226e5d 100644
--- a/letta/memory.py
+++ b/letta/memory.py
@@ -7,6 +7,7 @@
 from letta.llm_api.llm_api_tools import create
 from letta.prompts.gpt_summarize import SYSTEM as SUMMARY_PROMPT_SYSTEM
 from letta.schemas.agent import AgentState
+from letta.schemas.enums import MessageRole
 from letta.schemas.memory import Memory
 from letta.schemas.message import Message
 from letta.schemas.passage import Passage
@@ -50,7 +51,6 @@ def _format_summary_history(message_history: List[Message]):
 def summarize_messages(
     agent_state: AgentState,
     message_sequence_to_summarize: List[Message],
-    insert_acknowledgement_assistant_message: bool = True,
 ):
     """Summarize a message sequence using GPT"""
     # we need the context_window
@@ -70,13 +70,17 @@ def summarize_messages(
     dummy_user_id = agent_state.user_id
     dummy_agent_id = agent_state.id
     message_sequence = []
-    message_sequence.append(Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role="system", text=summary_prompt))
-    if insert_acknowledgement_assistant_message:
-        message_sequence.append(Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role="assistant", text=MESSAGE_SUMMARY_REQUEST_ACK))
-    message_sequence.append(Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role="user", text=summary_input))
+    message_sequence.append(Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role=MessageRole.system, text=summary_prompt))
+    message_sequence.append(
+        Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role=MessageRole.assistant, text=MESSAGE_SUMMARY_REQUEST_ACK)
+    )
+    message_sequence.append(Message(user_id=dummy_user_id, agent_id=dummy_agent_id, role=MessageRole.user, text=summary_input))
 
+    # TODO: We need to eventually have a separate LLM config for the summarizer LLM
+    llm_config_no_inner_thoughts = agent_state.llm_config.model_copy(deep=True)
+    llm_config_no_inner_thoughts.put_inner_thoughts_in_kwargs = False
     response = create(
-        llm_config=agent_state.llm_config,
+        llm_config=llm_config_no_inner_thoughts,
         user_id=agent_state.user_id,
         messages=message_sequence,
         stream=False,
diff --git a/tests/helpers/endpoints_helper.py b/tests/helpers/endpoints_helper.py
index 49bb0dc1f0..2f32e67d63 100644
--- a/tests/helpers/endpoints_helper.py
+++ b/tests/helpers/endpoints_helper.py
@@ -38,7 +38,7 @@
     FunctionCall,
     Message,
 )
-from letta.utils import get_human_text, get_persona_text
+from letta.utils import get_human_text, get_persona_text, json_dumps
 from tests.helpers.utils import cleanup
 
 # Generate uuid for agent name for this example
@@ -321,6 +321,40 @@ def check_agent_edit_core_memory(filename: str) -> LettaResponse:
     return response
 
 
+def check_agent_summarize_memory_simple(filename: str) -> LettaResponse:
+    """
+    Checks that the LLM is able to summarize its memory
+    """
+    # Set up client
+    client = create_client()
+    cleanup(client=client, agent_uuid=agent_uuid)
+
+    agent_state = setup_agent(client, filename)
+
+    # Send a couple messages
+    friend_name = "Shub"
+    client.user_message(agent_id=agent_state.id, message="Hey, how's it going? What do you think about this whole shindig")
+    client.user_message(agent_id=agent_state.id, message=f"By the way, my friend's name is {friend_name}!")
+    client.user_message(agent_id=agent_state.id, message="Does the number 42 ring a bell?")
+
+    # Summarize
+    agent = client.server._get_or_load_agent(agent_id=agent_state.id)
+    agent.summarize_messages_inplace()
+    print(f"Summarization succeeded: messages[1] = \n\n{json_dumps(agent.messages[1])}\n")
+
+    response = client.user_message(agent_id=agent_state.id, message="What is my friend's name?")
+    # Basic checks
+    assert_sanity_checks(response)
+
+    # Make sure my name was repeated back to me
+    assert_invoked_send_message_with_keyword(response.messages, friend_name)
+
+    # Make sure some inner monologue is present
+    assert_inner_monologue_is_present_and_valid(response.messages)
+
+    return response
+
+
 def run_embedding_endpoint(filename):
     # load JSON file
     config_data = json.load(open(filename, "r"))
diff --git a/tests/test_endpoints.py b/tests/test_endpoints.py
index 76494c0999..575ae13b85 100644
--- a/tests/test_endpoints.py
+++ b/tests/test_endpoints.py
@@ -7,6 +7,7 @@
     check_agent_archival_memory_retrieval,
     check_agent_edit_core_memory,
     check_agent_recall_chat_memory,
+    check_agent_summarize_memory_simple,
     check_agent_uses_external_tool,
     check_first_response_is_valid_for_llm_endpoint,
     check_response_contains_keyword,
@@ -58,14 +59,14 @@ def wrapper(*args, **kwargs):
 # ======================================================================================================================
 # OPENAI TESTS
 # ======================================================================================================================
-def test_openai_gpt_4_returns_valid_first_message():
+def test_openai_gpt_4o_returns_valid_first_message():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_first_response_is_valid_for_llm_endpoint(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_returns_keyword():
+def test_openai_gpt_4o_returns_keyword():
     keyword = "banana"
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_response_contains_keyword(filename, keyword=keyword)
@@ -73,41 +74,47 @@ def test_openai_gpt_4_returns_keyword():
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_uses_external_tool():
+def test_openai_gpt_4o_uses_external_tool():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_agent_uses_external_tool(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_recall_chat_memory():
+def test_openai_gpt_4o_recall_chat_memory():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_agent_recall_chat_memory(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_archival_memory_retrieval():
+def test_openai_gpt_4o_archival_memory_retrieval():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_agent_archival_memory_retrieval(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_archival_memory_insert():
+def test_openai_gpt_4o_archival_memory_insert():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_agent_archival_memory_insert(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
-def test_openai_gpt_4_edit_core_memory():
+def test_openai_gpt_4o_edit_core_memory():
     filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
     response = check_agent_edit_core_memory(filename)
     # Log out successful response
     print(f"Got successful response from client: \n\n{response}")
 
 
+def test_openai_gpt_4o_summarize_memory():
+    filename = os.path.join(llm_config_dir, "openai-gpt-4o.json")
+    response = check_agent_summarize_memory_simple(filename)
+    print(f"Got successful response from client: \n\n{response}")
+
+
 def test_embedding_endpoint_openai():
     filename = os.path.join(embedding_config_dir, "openai_embed.json")
     run_embedding_endpoint(filename)