Assistant - add support for gpt-5.1 model

Author: vertyco

assistant/abc.py

@@ -62,23 +62,23 @@ async def cut_text_by_tokens(self, text: str, conf: GuildSettings, user: Optiona
         raise NotImplementedError
 
     @abstractmethod
-    async def count_payload_tokens(self, messages: List[dict], model: str = "gpt-5") -> int:
+    async def count_payload_tokens(self, messages: List[dict], model: str = "gpt-5.1") -> int:
         raise NotImplementedError
 
     @abstractmethod
-    async def count_function_tokens(self, functions: List[dict], model: str = "gpt-5") -> int:
+    async def count_function_tokens(self, functions: List[dict], model: str = "gpt-5.1") -> int:
         raise NotImplementedError
 
     @abstractmethod
     async def count_tokens(self, text: str, model: str) -> int:
         raise NotImplementedError
 
     @abstractmethod
-    async def get_tokens(self, text: str, model: str = "gpt-5") -> list[int]:
+    async def get_tokens(self, text: str, model: str = "gpt-5.1") -> list[int]:
         raise NotImplementedError
 
     @abstractmethod
-    async def get_text(self, tokens: list, model: str = "gpt-5") -> str:
+    async def get_text(self, tokens: list, model: str = "gpt-5.1") -> str:
         raise NotImplementedError
 
     @abstractmethod

assistant/assistant.py

@@ -56,7 +56,7 @@ class Assistant(
     """
 
     __author__ = "[vertyco](https://github.com/vertyco/vrt-cogs)"
-    __version__ = "6.18.11"
+    __version__ = "6.18.12"
 
     def format_help_for_context(self, ctx):
         helpcmd = super().format_help_for_context(ctx)

assistant/commands/base.py

@@ -473,7 +473,7 @@ async def summarize_convo(
             response: ChatCompletionMessage = await self.request_response(
                 messages=payload,
                 conf=conf,
-                model_override="gpt-5",
+                model_override="gpt-5.1",
                 temperature_override=0.0,
             )
         except httpx.ReadTimeout:

assistant/common/api.py

@@ -81,8 +81,8 @@ async def request_response(
                 response_tokens = min(response_tokens, max_response_tokens)
 
         if model not in MODELS and self.db.endpoint_override is None:
-            log.error(f"This model is no longer supported: {model}. Switching to gpt-5")
-            model = "gpt-5"
+            log.error(f"This model is no longer supported: {model}. Switching to gpt-5.1")
+            model = "gpt-5.1"
             await self.save_conf()
 
         response: ChatCompletion = await request_chat_completion_raw(
@@ -132,7 +132,7 @@ async def request_embedding(self, text: str, conf: GuildSettings) -> List[float]
     # -------------------------------------------------------
     # -------------------------------------------------------
 
-    async def count_payload_tokens(self, messages: List[dict], model: str = "gpt-5") -> int:
+    async def count_payload_tokens(self, messages: List[dict], model: str = "gpt-5.1") -> int:
         if not messages:
             return 0
 
@@ -167,7 +167,7 @@ def _count_payload():
 
         return await asyncio.to_thread(_count_payload)
 
-    async def count_function_tokens(self, functions: List[dict], model: str = "gpt-5") -> int:
+    async def count_function_tokens(self, functions: List[dict], model: str = "gpt-5.1") -> int:
         # Initialize function settings to 0
         func_init = 0
         prop_init = 0
@@ -205,6 +205,8 @@ async def count_function_tokens(self, functions: List[dict], model: str = "gpt-5
             "gpt-5-mini-2025-04-16",
             "gpt-5-nano",
             "gpt-5-nano-2025-04-16",
+            "gpt-5.1",
+            "gpt-5.1-2025-11-13",
         ]:
             # Set function settings for the above models
             func_init = 7
@@ -273,7 +275,7 @@ def _count_tokens():
 
         return await asyncio.to_thread(_count_tokens)
 
-    async def get_tokens(self, text: str, model: str = "gpt-5") -> list[int]:
+    async def get_tokens(self, text: str, model: str = "gpt-5.1") -> list[int]:
         """Get token list from text"""
         if not text:
             log.debug("No text to get tokens from!")
@@ -358,7 +360,7 @@ async def cut_text_by_tokens(self, text: str, conf: GuildSettings, user: Optiona
         tokens = await self.get_tokens(text, conf.get_user_model(user))
         return await self.get_text(tokens[: self.get_max_tokens(conf, user)], conf.get_user_model(user))
 
-    async def get_text(self, tokens: list, model: str = "gpt-5") -> str:
+    async def get_text(self, tokens: list, model: str = "gpt-5.1") -> str:
         """Get text from token list"""
 
         def _get_encoding():

assistant/common/constants.py

@@ -43,6 +43,8 @@
     "gpt-5-mini-2025-08-07": 400000,
     "gpt-5-nano": 400000,
     "gpt-5-nano-2025-08-07": 400000,
+    "gpt-5.1": 400000,
+    "gpt-5.1-2025-11-13": 400000,
 }
 PRICES = {  # Price per 1k tokens
     "gpt-3.5-turbo": [0.001, 0.0015],
@@ -86,6 +88,8 @@
     "gpt-5-mini-2025-08-07": [0.00025, 0.002],
     "gpt-5-nano": [0.00005, 0.0004],
     "gpt-5-nano-2025-08-07": [0.00005, 0.0004],
+    "gpt-5.1": [0.00125, 0.010],
+    "gpt-5.1-2025-11-13": [0.00125, 0.010],
     "o1": [0.015, 0.06],
     "o1-2024-12-17": [0.015, 0.06],
     "o1-preview": [0.015, 0.06],
@@ -166,6 +170,8 @@
     "gpt-5-mini-2025-08-07",
     "gpt-5-nano",
     "gpt-5-nano-2025-08-07",
+    "gpt-5.1",
+    "gpt-5.1-2025-11-13",
 ]
 NO_DEVELOPER_ROLE = [  # Also doesnt support system messages
     "o1-mini",
@@ -204,6 +210,8 @@
     "gpt-5-mini-2025-08-07",
     "gpt-5-nano",
     "gpt-5-nano-2025-08-07",
+    "gpt-5.1",
+    "gpt-5.1-2025-11-13",
 ]
 SUPPORTS_TOOLS = [
     "gpt-3.5-turbo-1106",
@@ -239,6 +247,8 @@
     "gpt-5-mini-2025-08-07",
     "gpt-5-nano",
     "gpt-5-nano-2025-08-07",
+    "gpt-5.1",
+    "gpt-5.1-2025-11-13",
 ]
 READ_EXTENSIONS = [
     ".txt",

assistant/common/models.py

@@ -96,7 +96,7 @@ class GuildSettings(AssistantBaseModel):
     mention: bool = False
     mention_respond: bool = True
     enabled: bool = True  # Auto-reply channel
-    model: str = "gpt-5"
+    model: str = "gpt-5.1"
     embed_model: str = "text-embedding-3-small"  # Or text-embedding-3-large, text-embedding-ada-002
     collab_convos: bool = False
     reasoning_effort: str = "low"  # low, medium, high (or minimal for gpt-5)
@@ -106,7 +106,7 @@ class GuildSettings(AssistantBaseModel):
     auto_answer: bool = False  # Answer questions anywhere if one is detected and embedding is found for it
     auto_answer_threshold: float = 0.7  # 0.0 - 1.0  # Confidence threshold for auto-answer
     auto_answer_ignored_channels: t.List[int] = []  # Channel IDs to ignore auto-answer
-    auto_answer_model: str = "gpt-5"  # Model to use for auto-answer
+    auto_answer_model: str = "gpt-5.1"  # Model to use for auto-answer
 
     image_command: bool = True  # Allow image commands