diff --git a/autogen/agentchat/groupchat.py b/autogen/agentchat/groupchat.py
index 174f62f4e15..fd481632ed7 100644
--- a/autogen/agentchat/groupchat.py
+++ b/autogen/agentchat/groupchat.py
@@ -120,7 +120,6 @@ def __init__(
         max_consecutive_auto_reply: Optional[int] = sys.maxsize,
         human_input_mode: Optional[str] = "NEVER",
         system_message: Optional[str] = "Group chat manager.",
-        # seed: Optional[int] = 4,
         **kwargs,
     ):
         super().__init__(
@@ -136,8 +135,6 @@ def __init__(
         # Allow async chat if initiated using a_initiate_chat
         self.register_reply(Agent, GroupChatManager.a_run_chat, config=groupchat, reset_config=GroupChat.reset)
 
-        # self._random = random.Random(seed)
-
     def run_chat(
         self,
         messages: Optional[List[Dict]] = None,
diff --git a/autogen/code_utils.py b/autogen/code_utils.py
index 5e65a65f11c..616bcf6da57 100644
--- a/autogen/code_utils.py
+++ b/autogen/code_utils.py
@@ -509,11 +509,11 @@ def eval_function_completions(
 _FUNC_COMPLETION_PROMPT = "# Python 3{definition}"
 _FUNC_COMPLETION_STOP = ["\nclass", "\ndef", "\nif", "\nprint"]
 _IMPLEMENT_CONFIGS = [
-    {"model": FAST_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "temperature": 0, "seed": 0},
-    {"model": FAST_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 7, "seed": 0},
-    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "temperature": 0, "seed": 1},
-    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 2, "seed": 2},
-    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 1, "seed": 2},
+    {"model": FAST_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "temperature": 0, "cache_seed": 0},
+    {"model": FAST_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 7, "cache_seed": 0},
+    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "temperature": 0, "cache_seed": 1},
+    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 2, "cache_seed": 2},
+    {"model": DEFAULT_MODEL, "prompt": _FUNC_COMPLETION_PROMPT, "stop": _FUNC_COMPLETION_STOP, "n": 1, "cache_seed": 2},
 ]
 
 
diff --git a/autogen/oai/client.py b/autogen/oai/client.py
index 46748207bec..4b6a92c4dc2 100644
--- a/autogen/oai/client.py
+++ b/autogen/oai/client.py
@@ -31,7 +31,7 @@ class OpenAIWrapper:
     """A wrapper class for openai client."""
 
     cache_path_root: str = ".cache"
-    extra_kwargs = {"seed", "filter_func", "allow_format_str_template", "context", "api_version"}
+    extra_kwargs = {"cache_seed", "filter_func", "allow_format_str_template", "context", "api_version"}
     openai_kwargs = set(inspect.getfullargspec(OpenAI.__init__).kwonlyargs)
 
     def __init__(self, *, config_list: List[Dict] = None, **base_config):
@@ -191,8 +191,8 @@ def create(self, **config):
                 The actual prompt will be:
                 "Complete the following sentence: Today I feel".
                 More examples can be found at [templating](/docs/Use-Cases/enhanced_inference#templating).
-            - `seed` (int | None) for the cache. Default to 41.
-                An integer seed is useful when implementing "controlled randomness" for the completion.
+            - `cache_seed` (int | None) for the cache. Default to 41.
+                An integer cache_seed is useful when implementing "controlled randomness" for the completion.
                 None for no caching.
             - filter_func (Callable | None): A function that takes in the context and the response
                 and returns a boolean to indicate whether the response is valid. E.g.,
@@ -219,12 +219,12 @@ def yes_or_no_filter(context, response):
             self._process_for_azure(create_config, extra_kwargs, "extra")
             # construct the create params
             params = self._construct_create_params(create_config, extra_kwargs)
-            # get the seed, filter_func and context
-            seed = extra_kwargs.get("seed", 41)
+            # get the cache_seed, filter_func and context
+            cache_seed = extra_kwargs.get("cache_seed", 41)
             filter_func = extra_kwargs.get("filter_func")
             context = extra_kwargs.get("context")
-            with diskcache.Cache(f"{self.cache_path_root}/{seed}") as cache:
-                if seed is not None:
+            with diskcache.Cache(f"{self.cache_path_root}/{cache_seed}") as cache:
+                if cache_seed is not None:
                     # Try to get the response from cache
                     key = get_key(params)
                     response = cache.get(key, None)
@@ -245,7 +245,7 @@ def yes_or_no_filter(context, response):
                     if i == last:
                         raise
                 else:
-                    if seed is not None:
+                    if cache_seed is not None:
                         # Cache the response
                         cache.set(key, response)
                     return response
diff --git a/autogen/oai/completion.py b/autogen/oai/completion.py
index 88d53bca4c0..2fe95e8f60c 100644
--- a/autogen/oai/completion.py
+++ b/autogen/oai/completion.py
@@ -109,8 +109,8 @@ class Completion(openai_Completion):
         "prompt": "{prompt}",
     }
 
-    seed = 41
-    cache_path = f".cache/{seed}"
+    cache_seed = 41
+    cache_path = f".cache/{cache_seed}"
     # retry after this many seconds
     retry_wait_time = 10
     # fail a request after hitting RateLimitError for this many seconds
@@ -134,7 +134,7 @@ def set_cache(cls, seed: Optional[int] = 41, cache_path_root: Optional[str] = ".
             cache_path (str, Optional): The root path for the cache.
                 The complete cache path will be {cache_path}/{seed}.
         """
-        cls.seed = seed
+        cls.cache_seed = seed
         cls.cache_path = f"{cache_path_root}/{seed}"
 
     @classmethod
@@ -145,7 +145,7 @@ def clear_cache(cls, seed: Optional[int] = None, cache_path_root: Optional[str]
             seed (int, Optional): The integer identifier for the pseudo seed.
                 If omitted, all caches under cache_path_root will be cleared.
             cache_path (str, Optional): The root path for the cache.
-                The complete cache path will be {cache_path}/{seed}.
+                The complete cache path will be {cache_path}/{cache_seed}.
         """
         if seed is None:
             shutil.rmtree(cache_path_root, ignore_errors=True)
@@ -773,7 +773,7 @@ def yes_or_no_filter(context, config, response):
                 Besides the parameters for the openai API call, it can also contain:
                 - `max_retry_period` (int): the total time (in seconds) allowed for retrying failed requests.
                 - `retry_wait_time` (int): the time interval to wait (in seconds) before retrying a failed request.
-                - `seed` (int) for the cache. This is useful when implementing "controlled randomness" for the completion.
+                - `cache_seed` (int) for the cache. This is useful when implementing "controlled randomness" for the completion.
 
         Returns:
             Responses from OpenAI API, with additional fields.
@@ -831,11 +831,11 @@ def yes_or_no_filter(context, config, response):
             return cls._get_response(
                 params, raise_on_ratelimit_or_timeout=raise_on_ratelimit_or_timeout, use_cache=False
             )
-        seed = cls.seed
-        if "seed" in params:
-            cls.set_cache(params.pop("seed"))
+        cache_seed = cls.cache_seed
+        if "cache_seed" in params:
+            cls.set_cache(params.pop("cache_seed"))
         with diskcache.Cache(cls.cache_path) as cls._cache:
-            cls.set_cache(seed)
+            cls.set_cache(cache_seed)
             return cls._get_response(params, raise_on_ratelimit_or_timeout=raise_on_ratelimit_or_timeout)
 
     @classmethod
diff --git a/autogen/version.py b/autogen/version.py
index 24374a4e9d7..16bf2ffe042 100644
--- a/autogen/version.py
+++ b/autogen/version.py
@@ -1 +1 @@
-__version__ = "0.2.0b3"
+__version__ = "0.2.0b4"
diff --git a/notebook/agentchat_RetrieveChat.ipynb b/notebook/agentchat_RetrieveChat.ipynb
index e0aa629cb75..faee4cf2bf5 100644
--- a/notebook/agentchat_RetrieveChat.ipynb
+++ b/notebook/agentchat_RetrieveChat.ipynb
@@ -185,7 +185,7 @@
     "    system_message=\"You are a helpful assistant.\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": config_list,\n",
     "    },\n",
     ")\n",
diff --git a/notebook/agentchat_auto_feedback_from_code_execution.ipynb b/notebook/agentchat_auto_feedback_from_code_execution.ipynb
index 21543ffc208..1f29c6d0aa5 100644
--- a/notebook/agentchat_auto_feedback_from_code_execution.ipynb
+++ b/notebook/agentchat_auto_feedback_from_code_execution.ipynb
@@ -45,7 +45,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -330,7 +330,7 @@
     "assistant = autogen.AssistantAgent(\n",
     "    name=\"assistant\",\n",
     "    llm_config={\n",
-    "        \"seed\": 42,  # seed for caching and reproducibility\n",
+    "        \"cache_seed\": 42,  # seed for caching and reproducibility\n",
     "        \"config_list\": config_list,  # a list of OpenAI API configurations\n",
     "        \"temperature\": 0,  # temperature for sampling\n",
     "    },  # configuration for autogen's enhanced inference API which is compatible with OpenAI API\n",
@@ -806,7 +806,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.17"
+   "version": "3.11.4"
   },
   "vscode": {
    "interpreter": {
diff --git a/notebook/agentchat_chess.ipynb b/notebook/agentchat_chess.ipynb
index 06bee1d6591..6b2a64a2a37 100644
--- a/notebook/agentchat_chess.ipynb
+++ b/notebook/agentchat_chess.ipynb
@@ -35,7 +35,7 @@
    "outputs": [],
    "source": [
     "%%capture --no-stderr\n",
-    "# %pip install \"pyautogen~=0.2.0b2\"\n",
+    "# %pip install \"pyautogen~=0.2.0b4\"\n",
     "%pip install chess -U"
    ]
   },
@@ -292,13 +292,13 @@
     "    color=\"black\",\n",
     "    board_agent=board_agent,\n",
     "    max_turns=max_turn,\n",
-    "    llm_config={\"temperature\": 0.5, \"seed\": 1, \"config_list\": config_list_gpt4},\n",
+    "    llm_config={\"temperature\": 0.5, \"cache_seed\": 1, \"config_list\": config_list_gpt4},\n",
     ")\n",
     "player_white = ChessPlayerAgent(\n",
     "    color=\"white\",\n",
     "    board_agent=board_agent,\n",
     "    max_turns=max_turn,\n",
-    "    llm_config={\"temperature\": 0.5, \"seed\": 2, \"config_list\": config_list_gpt4},\n",
+    "    llm_config={\"temperature\": 0.5, \"cache_seed\": 2, \"config_list\": config_list_gpt4},\n",
     ")"
    ]
   },
diff --git a/notebook/agentchat_groupchat.ipynb b/notebook/agentchat_groupchat.ipynb
index fccd7449921..042aebf4b6f 100644
--- a/notebook/agentchat_groupchat.ipynb
+++ b/notebook/agentchat_groupchat.ipynb
@@ -35,7 +35,7 @@
    "outputs": [],
    "source": [
     "%%capture --no-stderr\n",
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -124,7 +124,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42}\n",
+    "llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": 42}\n",
     "user_proxy = autogen.UserProxyAgent(\n",
     "   name=\"User_proxy\",\n",
     "   system_message=\"A human admin.\",\n",
diff --git a/notebook/agentchat_groupchat_research.ipynb b/notebook/agentchat_groupchat_research.ipynb
index abe408320ae..b01e35f8692 100644
--- a/notebook/agentchat_groupchat_research.ipynb
+++ b/notebook/agentchat_groupchat_research.ipynb
@@ -33,7 +33,7 @@
    "outputs": [],
    "source": [
     "%%capture --no-stderr\n",
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -111,7 +111,7 @@
    "outputs": [],
    "source": [
     "gpt4_config = {\n",
-    "    \"seed\": 42,  # change the seed for different trials\n",
+    "    \"cache_seed\": 42,  # change the cache_seed for different trials\n",
     "    \"temperature\": 0,\n",
     "    \"config_list\": config_list_gpt4,\n",
     "    \"timeout\": 120,\n",
diff --git a/notebook/agentchat_groupchat_vis.ipynb b/notebook/agentchat_groupchat_vis.ipynb
index 499e6cee4fc..0a7b984a551 100644
--- a/notebook/agentchat_groupchat_vis.ipynb
+++ b/notebook/agentchat_groupchat_vis.ipynb
@@ -33,7 +33,7 @@
    "outputs": [],
    "source": [
     "%%capture --no-stderr\n",
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -132,7 +132,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42}\n",
+    "llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": 42}\n",
     "user_proxy = autogen.UserProxyAgent(\n",
     "   name=\"User_proxy\",\n",
     "   system_message=\"A human admin.\",\n",
diff --git a/notebook/agentchat_hierarchy_flow_using_select_speaker.ipynb b/notebook/agentchat_hierarchy_flow_using_select_speaker.ipynb
index bf69277a233..d03157b6dc6 100644
--- a/notebook/agentchat_hierarchy_flow_using_select_speaker.ipynb
+++ b/notebook/agentchat_hierarchy_flow_using_select_speaker.ipynb
@@ -50,7 +50,7 @@
    "outputs": [],
    "source": [
     "%%capture --no-stderr\n",
-    "# %pip install pyautogen~=0.1.0"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -163,7 +163,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42}"
+    "llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": 42}"
    ]
   },
   {
@@ -359,7 +359,7 @@
     "\n",
     "\n",
     "# Create the manager\n",
-    "llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42, \"use_cache\":False}  # use_cache is False because we want to observe if there is any communication pattern difference if we reran the group chat.\n",
+    "llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": None}  # cache_seed is None because we want to observe if there is any communication pattern difference if we reran the group chat.\n",
     "manager = autogen.GroupChatManager(groupchat=group_chat, llm_config=llm_config)\n",
     "\n",
     "\n",
diff --git a/notebook/agentchat_human_feedback.ipynb b/notebook/agentchat_human_feedback.ipynb
index 031d4901e9c..93b3afdf62d 100644
--- a/notebook/agentchat_human_feedback.ipynb
+++ b/notebook/agentchat_human_feedback.ipynb
@@ -45,7 +45,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -123,7 +123,7 @@
     "assistant = autogen.AssistantAgent(\n",
     "    name=\"assistant\",\n",
     "    llm_config={\n",
-    "        \"seed\": 41,\n",
+    "        \"cache_seed\": 41,\n",
     "        \"config_list\": config_list,\n",
     "    }\n",
     ")\n",
diff --git a/notebook/agentchat_lmm_gpt-4v.ipynb b/notebook/agentchat_lmm_gpt-4v.ipynb
index 50e018076bb..d3ece5c66dd 100644
--- a/notebook/agentchat_lmm_gpt-4v.ipynb
+++ b/notebook/agentchat_lmm_gpt-4v.ipynb
@@ -15,7 +15,7 @@
    "source": [
     "### Before everything starts, install AutoGen with the `lmm` option\n",
     "```bash\n",
-    "pip install pyautogen[lmm]\n",
+    "pip install \"pyautogen[lmm]~=0.2.0b4\"\n",
     "```"
    ]
   },
@@ -85,7 +85,7 @@
     "    },\n",
     ")\n",
     "\n",
-    "gpt4_llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42}"
+    "gpt4_llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": 42}"
    ]
   },
   {
@@ -699,7 +699,7 @@
     "#     },\n",
     "# )\n",
     "\n",
-    "# gpt35_llm_config = {\"config_list\": config_list_gpt35, \"seed\": 42}\n",
+    "# gpt35_llm_config = {\"config_list\": config_list_gpt35, \"cache_seed\": 42}\n",
     "\n",
     "\n",
     "creator = FigureCreator(\n",
diff --git a/notebook/agentchat_lmm_llava.ipynb b/notebook/agentchat_lmm_llava.ipynb
index 35a51f45ad9..e86a1007e68 100644
--- a/notebook/agentchat_lmm_llava.ipynb
+++ b/notebook/agentchat_lmm_llava.ipynb
@@ -5,9 +5,9 @@
    "id": "2c75da30",
    "metadata": {},
    "source": [
-    "# Agent Chat with Multimodal Models\n",
+    "# Agent Chat with Multimodal Models: LLaVA\n",
     "\n",
-    "We use **LLaVA** as an example for the multimodal feature.  More information about LLaVA can be found in their [GitHub page](https://github.com/haotian-liu/LLaVA)\n",
+    "This notebook uses **LLaVA** as an example for the multimodal feature. More information about LLaVA can be found in their [GitHub page](https://github.com/haotian-liu/LLaVA)\n",
     "\n",
     "\n",
     "This notebook contains the following information and examples:\n",
@@ -26,7 +26,7 @@
    "source": [
     "### Before everything starts, install AutoGen with the `lmm` option\n",
     "```bash\n",
-    "pip install pyautogen[lmm]\n",
+    "pip install \"pyautogen[lmm]~=0.2.0b4\"\n",
     "```"
    ]
   },
@@ -848,7 +848,7 @@
     "    },\n",
     ")\n",
     "\n",
-    "gpt4_llm_config = {\"config_list\": config_list_gpt4, \"seed\": 42}\n",
+    "gpt4_llm_config = {\"config_list\": config_list_gpt4, \"cache_seed\": 42}\n",
     "\n",
     "# config_list_gpt35 = autogen.config_list_from_json(\n",
     "#     \"OAI_CONFIG_LIST\",\n",
@@ -857,7 +857,7 @@
     "#     },\n",
     "# )\n",
     "\n",
-    "# gpt35_llm_config = {\"config_list\": config_list_gpt35, \"seed\": 42}\n",
+    "# gpt35_llm_config = {\"config_list\": config_list_gpt35, \"cache_seed\": 42}\n",
     "\n",
     "\n",
     "creator = FigureCreator(\n",
diff --git a/notebook/agentchat_planning.ipynb b/notebook/agentchat_planning.ipynb
index 70f5ad150f5..b1a6789560e 100644
--- a/notebook/agentchat_planning.ipynb
+++ b/notebook/agentchat_planning.ipynb
@@ -45,7 +45,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2 docker"
+    "# %pip install pyautogen~=0.2.0b4 docker"
    ]
   },
   {
@@ -156,7 +156,7 @@
     "    llm_config={\n",
     "        \"temperature\": 0,\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": config_list,\n",
     "        \"functions\": [\n",
     "            {\n",
diff --git a/notebook/agentchat_stream.ipynb b/notebook/agentchat_stream.ipynb
index d3f485305c8..3ea61fcc359 100644
--- a/notebook/agentchat_stream.ipynb
+++ b/notebook/agentchat_stream.ipynb
@@ -45,7 +45,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -209,7 +209,7 @@
     "    name=\"assistant\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 41,\n",
+    "        \"cache_seed\": 41,\n",
     "        \"config_list\": config_list,\n",
     "        \"temperature\": 0,\n",
     "    },\n",
diff --git a/notebook/agentchat_two_users.ipynb b/notebook/agentchat_two_users.ipynb
index 0f29374f292..b170d2ec63f 100644
--- a/notebook/agentchat_two_users.ipynb
+++ b/notebook/agentchat_two_users.ipynb
@@ -44,7 +44,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2"
+    "# %pip install pyautogen~=0.2.0b4"
    ]
   },
   {
@@ -161,7 +161,7 @@
     "    system_message=\"You are a helpful assistant. Reply TERMINATE when the task is done.\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": config_list,\n",
     "        \"temperature\": 0,\n",
     "        \"functions\": [\n",
diff --git a/notebook/agentchat_web_info.ipynb b/notebook/agentchat_web_info.ipynb
index 849b4f4c2f5..2dc5339544c 100644
--- a/notebook/agentchat_web_info.ipynb
+++ b/notebook/agentchat_web_info.ipynb
@@ -49,7 +49,7 @@
    },
    "outputs": [],
    "source": [
-    "# %pip install pyautogen~=0.2.0b2 docker"
+    "# %pip install pyautogen~=0.2.0b4 docker"
    ]
   },
   {
@@ -79,7 +79,7 @@
     "\n",
     "llm_config={\n",
     "    \"timeout\": 600,\n",
-    "    \"seed\": 42,\n",
+    "    \"cache_seed\": 42,\n",
     "    \"config_list\": config_list,\n",
     "    \"temperature\": 0,\n",
     "}"
diff --git a/notebook/oai_openai_utils.ipynb b/notebook/oai_openai_utils.ipynb
index 24973e3d908..3c020b4ab1a 100644
--- a/notebook/oai_openai_utils.ipynb
+++ b/notebook/oai_openai_utils.ipynb
@@ -39,7 +39,7 @@
     "    name=\"assistant\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": config_list,\n",
     "        \"temperature\": 0,\n",
     "    },\n",
@@ -60,7 +60,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# ! pip install pyautogen"
+    "# %pip install \"pyautogen~=0.2.0b4\""
    ]
   },
   {
@@ -227,7 +227,7 @@
     "    name=\"3.5-assistant\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": cheap_config_list,\n",
     "        \"temperature\": 0,\n",
     "    },\n",
@@ -238,7 +238,7 @@
     "    name=\"4-assistant\",\n",
     "    llm_config={\n",
     "        \"timeout\": 600,\n",
-    "        \"seed\": 42,\n",
+    "        \"cache_seed\": 42,\n",
     "        \"config_list\": costly_config_list,\n",
     "        \"temperature\": 0,\n",
     "    },\n",
diff --git a/test/agentchat/test_assistant_agent.py b/test/agentchat/test_assistant_agent.py
index 61510a44f9c..361c71f56fc 100644
--- a/test/agentchat/test_assistant_agent.py
+++ b/test/agentchat/test_assistant_agent.py
@@ -31,7 +31,7 @@ def test_ai_user_proxy_agent():
         system_message="You are a helpful assistant.",
         llm_config={
             "timeout": 600,
-            "seed": 42,
+            "cache_seed": 42,
             "config_list": config_list,
         },
     )
@@ -78,7 +78,7 @@ def test_gpt35(human_input_mode="NEVER", max_consecutive_auto_reply=5):
         },
     )
     llm_config = {
-        "seed": 42,
+        "cache_seed": 42,
         "config_list": config_list,
         "max_tokens": 1024,
     }
@@ -125,7 +125,7 @@ def test_create_execute_script(human_input_mode="NEVER", max_consecutive_auto_re
     # autogen.ChatCompletion.start_logging(conversations)
     llm_config = {
         "timeout": 600,
-        "seed": 42,
+        "cache_seed": 42,
         "config_list": config_list,
     }
     assistant = AssistantAgent(
diff --git a/test/agentchat/test_async.py b/test/agentchat/test_async.py
index ba332b25379..76176131015 100644
--- a/test/agentchat/test_async.py
+++ b/test/agentchat/test_async.py
@@ -71,7 +71,7 @@ async def add_stock_price_data():
         name="assistant",
         llm_config={
             "timeout": 600,
-            "seed": 41,
+            "cache_seed": 41,
             "config_list": config_list,
             "temperature": 0,
         },
diff --git a/test/agentchat/test_async_get_human_input.py b/test/agentchat/test_async_get_human_input.py
index fc8a8e31504..2daca19f2d9 100644
--- a/test/agentchat/test_async_get_human_input.py
+++ b/test/agentchat/test_async_get_human_input.py
@@ -16,7 +16,7 @@ async def test_async_get_human_input():
     assistant = autogen.AssistantAgent(
         name="assistant",
         max_consecutive_auto_reply=2,
-        llm_config={"timeout": 600, "seed": 41, "config_list": config_list, "temperature": 0},
+        llm_config={"timeout": 600, "cache_seed": 41, "config_list": config_list, "temperature": 0},
     )
 
     user_proxy = autogen.UserProxyAgent(name="user", human_input_mode="ALWAYS", code_execution_config=False)
diff --git a/test/agentchat/test_function_call_groupchat.py b/test/agentchat/test_function_call_groupchat.py
index fefb620ad1c..48414903562 100644
--- a/test/agentchat/test_function_call_groupchat.py
+++ b/test/agentchat/test_function_call_groupchat.py
@@ -30,7 +30,7 @@ def get_random_number():
     )
     llm_config = {
         "config_list": config_list_gpt4,
-        "seed": 42,
+        "cache_seed": 42,
         "functions": [
             {
                 "name": "get_random_number",
diff --git a/test/agentchat/test_math_user_proxy_agent.py b/test/agentchat/test_math_user_proxy_agent.py
index 2975e8d0f16..565aa80eb36 100644
--- a/test/agentchat/test_math_user_proxy_agent.py
+++ b/test/agentchat/test_math_user_proxy_agent.py
@@ -38,7 +38,7 @@ def test_math_user_proxy_agent():
         system_message="You are a helpful assistant.",
         llm_config={
             "timeout": 600,
-            "seed": 42,
+            "cache_seed": 42,
             "config_list": config_list,
         },
     )
diff --git a/test/agentchat/test_teachable_agent.py b/test/agentchat/test_teachable_agent.py
index 5a5be0379a1..712ed553f37 100644
--- a/test/agentchat/test_teachable_agent.py
+++ b/test/agentchat/test_teachable_agent.py
@@ -25,7 +25,7 @@ def colored(x, *args, **kwargs):
 
 assert_on_error = False  # GPT-4 nearly always succeeds on these unit tests, but GPT-3.5 is a bit less reliable.
 recall_threshold = 1.5  # Higher numbers allow more (but less relevant) memos to be recalled.
-seed = None
+cache_seed = None
 # If int, cached LLM calls will be skipped and responses pulled from cache. None exposes LLM non-determinism.
 
 # Specify the model to use by uncommenting one of the following lines.
@@ -43,7 +43,7 @@ def create_teachable_agent(reset_db=False, verbosity=0):
     config_list = config_list_from_json(env_or_file=OAI_CONFIG_LIST, filter_dict=filter_dict, file_location=KEY_LOC)
     teachable_agent = TeachableAgent(
         name="teachableagent",
-        llm_config={"config_list": config_list, "timeout": 120, "seed": seed},
+        llm_config={"config_list": config_list, "timeout": 120, "cache_seed": cache_seed},
         teach_config={
             "verbosity": verbosity,
             "reset_db": reset_db,
diff --git a/test/oai/test_client.py b/test/oai/test_client.py
index c74332588b6..83e01bceddf 100644
--- a/test/oai/test_client.py
+++ b/test/oai/test_client.py
@@ -18,7 +18,11 @@ def test_aoai_chat_completion():
         filter_dict={"api_type": ["azure"], "model": ["gpt-3.5-turbo"]},
     )
     client = OpenAIWrapper(config_list=config_list)
-    response = client.create(messages=[{"role": "user", "content": "2+2="}])
+    # for config in config_list:
+    #     print(config)
+    #     client = OpenAIWrapper(**config)
+    #     response = client.create(messages=[{"role": "user", "content": "2+2="}], cache_seed=None)
+    response = client.create(messages=[{"role": "user", "content": "2+2="}], cache_seed=None)
     print(response)
     print(client.extract_text_or_function_call(response))
 
diff --git a/website/docs/Installation.md b/website/docs/Installation.md
index d03e50c1b50..4100f1ffc23 100644
--- a/website/docs/Installation.md
+++ b/website/docs/Installation.md
@@ -65,7 +65,10 @@ print(client.extract_text_or_function_call(response))
 ```
 - Inference parameter tuning and inference logging features are currently unavailable in `OpenAIWrapper`. Logging will be added in a future release.
 Inference parameter tuning can be done via [`flaml.tune`](https://microsoft.github.io/FLAML/docs/Use-Cases/Tune-User-Defined-Function).
-- `use_cache` is removed as a kwarg in `OpenAIWrapper.create()` for being automatically decided by `seed`: int | None.
+- `seed` in autogen is renamed into `cache_seed` to accommodate the newly added `seed` param in openai chat completion api. `use_cache` is removed as a kwarg in `OpenAIWrapper.create()` for being automatically decided by `cache_seed`: int | None. The difference between autogen's `cache_seed` and openai's `seed` is that:
+    * autogen uses local disk cache to guarantee the exactly same output is produced for the same input and when cache is hit, no openai api call will be made.
+    * openai's `seed` is a best-effort deterministic sampling with no guarantee of determinism. When using openai's `seed` with `cache_seed` set to None, even for the same input, an openai api call will be made and there is no guarantee for getting exactly the same output.
+
 
 ### Optional Dependencies
 - #### docker
diff --git a/website/docs/Use-Cases/enhanced_inference.md b/website/docs/Use-Cases/enhanced_inference.md
index 5b5a1e81101..ffd4fd60f78 100644
--- a/website/docs/Use-Cases/enhanced_inference.md
+++ b/website/docs/Use-Cases/enhanced_inference.md
@@ -137,19 +137,23 @@ For local LLMs, one can spin up an endpoint using a package like [FastChat](http
 
 ## Caching
 
-API call results are cached locally and reused when the same request is issued. This is useful when repeating or continuing experiments for reproducibility and cost saving. It still allows controlled randomness by setting the "seed" specified in `OpenAIWrapper.create()` or the constructor of `OpenAIWrapper`.
+API call results are cached locally and reused when the same request is issued. This is useful when repeating or continuing experiments for reproducibility and cost saving. It still allows controlled randomness by setting the "cache_seed" specified in `OpenAIWrapper.create()` or the constructor of `OpenAIWrapper`.
 
 ```python
-client = OpenAIWrapper(seed=...)
+client = OpenAIWrapper(cache_seed=...)
 client.create(...)
 ```
 
 ```python
 client = OpenAIWrapper()
-client.create(seed=..., ...)
+client.create(cache_seed=..., ...)
 ```
 
-Caching is enabled by default with seed 41. To disable it please set `seed` to None.
+Caching is enabled by default with cache_seed 41. To disable it please set `cache_seed` to None.
+
+_NOTE_. openai v1.1 introduces a new param `seed`. The difference between autogen's `cache_seed` and openai's `seed` is that:
+* autogen uses local disk cache to guarantee the exactly same output is produced for the same input and when cache is hit, no openai api call will be made.
+* openai's `seed` is a best-effort deterministic sampling with no guarantee of determinism. When using openai's `seed` with `cache_seed` set to None, even for the same input, an openai api call will be made and there is no guarantee for getting exactly the same output.
 
 ## Error handling