unique-toolkit 2026.30.0.dev10__tar.gz → 2026.30.0.dev12__tar.gz

{unique_toolkit-2026.30.0.dev10 → unique_toolkit-2026.30.0.dev12}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-toolkit
-Version: 2026.30.0.dev10
+Version: 2026.30.0.dev12
 Summary: 
 Author: Cedric Klinkert, Martin Fadler, Sadique Sheik, Fabian Schläpfer, Pascal Hauri, Ejiroghene Oki-Peter, Rami Azouz, Ricardo Visini, Sanzio Monti
 Author-email: Cedric Klinkert <cedric.klinkert@unique.ch>, Martin Fadler <martin.fadler@unique.ch>, Sadique Sheik <sadique@unique.ch>, Fabian Schläpfer <fabian@unique.ch>, Pascal Hauri <pascal@unique.ch>, Ejiroghene Oki-Peter <oki-peter.ejiroghene@unique.ai>, Rami Azouz <rami.ext@unique.ch>, Ricardo Visini <ricardo.visini@unique.ch>, Sanzio Monti <sanzio.monti@unique.ai>
@@ -16,7 +16,7 @@ Requires-Dist: sseclient>=0.0.27,<0.0.28
 Requires-Dist: openai>=1.105.0,<3
 Requires-Dist: platformdirs>=4.0.0,<5
 Requires-Dist: pillow>=12.1.1
-Requires-Dist: unique-sdk>=2026.30.0.dev3,<2026.30.0rc0
+Requires-Dist: unique-sdk>=2026.30.0.dev5,<2026.30.0rc0
 Requires-Dist: jambo>=0.1.2,<0.2
 Requires-Dist: jinja2>=3.1.6,<4
 Requires-Dist: docxtpl>=0.20.1,<0.21

{unique_toolkit-2026.30.0.dev10 → unique_toolkit-2026.30.0.dev12}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "unique_toolkit"
-version = "2026.30.0.dev10"
+version = "2026.30.0.dev12"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12"
@@ -28,7 +28,7 @@ dependencies = [
     "openai>=1.105.0,<3",
     "platformdirs>=4.0.0,<5",
     "pillow>=12.1.1",
-    "unique-sdk>=2026.30.0.dev3,<2026.30.0rc0",
+    "unique-sdk>=2026.30.0.dev5,<2026.30.0rc0",
     "jambo>=0.1.2,<0.2",
     "jinja2>=3.1.6,<4",
     "docxtpl>=0.20.1,<0.21",

{unique_toolkit-2026.30.0.dev10 → unique_toolkit-2026.30.0.dev12}/unique_toolkit/chat/functions.py

@@ -66,6 +66,7 @@ def modify_message(
     debug_info: dict[str, Any] | None = None,
     message_id: str | None = None,
     set_completed_at: bool = False,
+    set_stopped_streaming_at: bool = False,
     segment_kind: str | None = None,
 ) -> ChatMessage:
     """Modifies a chat message synchronously.
@@ -84,6 +85,7 @@ def modify_message(
         references (list[ContentReference]): list of ContentReference objects. Defaults to None.
         debug_info (dict[str, Any]], optional): Debug information. Defaults to None.
         set_completed_at (bool, optional): Whether to set the completedAt field with the current date time. Defaults to False.
+        set_stopped_streaming_at (bool, optional): Whether to set the stoppedStreamingAt field with the current date time. Use this to signal that model streaming has finished while post-processing (e.g. evaluations) is still running, so clients can unblock user input before completedAt is set. Defaults to False.
         segment_kind (str, optional): The segment kind (e.g. "PROCESS", "PREFACE", "ELICITATION", "ANSWER") to relabel the segment to in place. Defaults to None (kind unchanged).
 
     Returns:
@@ -108,6 +110,7 @@ def modify_message(
             debug_info=debug_info,
             message_id=message_id,
             set_completed_at=set_completed_at,
+            set_stopped_streaming_at=set_stopped_streaming_at,
             segment_kind=segment_kind,
         )
         message = unique_sdk.Message.modify(**params)
@@ -131,6 +134,7 @@ async def modify_message_async(
     debug_info: dict[str, Any] | None = None,
     message_id: str | None = None,
     set_completed_at: bool = False,
+    set_stopped_streaming_at: bool = False,
     segment_kind: str | None = None,
 ) -> ChatMessage:
     """Modifies a chat message asynchronously.
@@ -149,6 +153,7 @@ async def modify_message_async(
         references (list[ContentReference]): list of ContentReference objects. Defaults to None.
         debug_info (dict[str, Any]], optional): Debug information. Defaults to None.
         set_completed_at (bool, optional): Whether to set the completedAt field with the current date time. Defaults to False.
+        set_stopped_streaming_at (bool, optional): Whether to set the stoppedStreamingAt field with the current date time. Use this to signal that model streaming has finished while post-processing (e.g. evaluations) is still running, so clients can unblock user input before completedAt is set. Defaults to False.
         segment_kind (str, optional): The segment kind (e.g. "PROCESS", "PREFACE", "ELICITATION", "ANSWER") to relabel the segment to in place. Defaults to None (kind unchanged).
 
     Returns:
@@ -173,6 +178,7 @@ async def modify_message_async(
             debug_info=debug_info,
             message_id=message_id,
             set_completed_at=set_completed_at,
+            set_stopped_streaming_at=set_stopped_streaming_at,
             segment_kind=segment_kind,
         )
         message = await unique_sdk.Message.modify_async(**params)
@@ -231,9 +237,11 @@ def _construct_message_modify_params(
     debug_info: dict[str, Any] | None = None,
     message_id: str | None = None,
     set_completed_at: bool = False,
+    set_stopped_streaming_at: bool = False,
     segment_kind: str | None = None,
 ) -> dict[str, Any]:
     completed_at_datetime = None
+    stopped_streaming_at_datetime = None
 
     if message_id:
         # Message ID specified. No need to guess
@@ -249,6 +257,9 @@ def _construct_message_modify_params(
     if set_completed_at:
         completed_at_datetime = _time_utils.get_datetime_now()
 
+    if set_stopped_streaming_at:
+        stopped_streaming_at_datetime = _time_utils.get_datetime_now()
+
     params: dict[str, Any] = {
         "user_id": user_id,
         "company_id": company_id,
@@ -259,6 +270,7 @@ def _construct_message_modify_params(
         "references": map_references(references) if references else [],
         "debugInfo": debug_info,
         "completedAt": completed_at_datetime,
+        "stoppedStreamingAt": stopped_streaming_at_datetime,
     }
     if segment_kind is not None:
         params["segmentKind"] = segment_kind

{unique_toolkit-2026.30.0.dev10 → unique_toolkit-2026.30.0.dev12}/unique_toolkit/language_model/infos.py

@@ -61,6 +61,9 @@ class LanguageModelName(StrEnum):
     AZURE_GPT_54_PRO_2026_0305 = "AZURE_GPT_54_PRO_2026_0305"
     AZURE_GPT_55_2026_0424 = "AZURE_GPT_55_2026_0424"
     AZURE_GPT_55_PRO_2026_0424 = "AZURE_GPT_55_PRO_2026_0424"
+    AZURE_GPT_56_SOL_2026_0709 = "AZURE_GPT_56_SOL_2026_0709"
+    AZURE_GPT_56_TERRA_2026_0709 = "AZURE_GPT_56_TERRA_2026_0709"
+    AZURE_GPT_56_LUNA_2026_0709 = "AZURE_GPT_56_LUNA_2026_0709"
     AZURE_GPT_4o_2024_0513 = "AZURE_GPT_4o_2024_0513"
     AZURE_GPT_4o_2024_0806 = "AZURE_GPT_4o_2024_0806"
     AZURE_GPT_4o_2024_1120 = "AZURE_GPT_4o_2024_1120"
@@ -117,6 +120,9 @@ class LanguageModelName(StrEnum):
     LITELLM_OPENAI_GPT_54_THINKING = "litellm:openai-gpt-5-4-thinking"
     LITELLM_OPENAI_GPT_55 = "litellm:openai-gpt-5-5"
     LITELLM_OPENAI_GPT_55_PRO = "litellm:openai-gpt-5-5-pro"
+    LITELLM_OPENAI_GPT_56_SOL = "litellm:openai-gpt-5-6-sol"
+    LITELLM_OPENAI_GPT_56_TERRA = "litellm:openai-gpt-5-6-terra"
+    LITELLM_OPENAI_GPT_56_LUNA = "litellm:openai-gpt-5-6-luna"
     LITELLM_OPENAI_O1 = "litellm:openai-o1"
     LITELLM_OPENAI_O3 = "litellm:openai-o3"
     LITELLM_OPENAI_O3_DEEP_RESEARCH = "litellm:openai-o3-deep-research"
@@ -201,6 +207,9 @@ def get_encoder_name(model_name: LanguageModelName) -> EncoderName:
             | LMN.AZURE_GPT_54_PRO_2026_0305
             | LMN.AZURE_GPT_55_2026_0424
             | LMN.AZURE_GPT_55_PRO_2026_0424
+            | LMN.AZURE_GPT_56_SOL_2026_0709
+            | LMN.AZURE_GPT_56_TERRA_2026_0709
+            | LMN.AZURE_GPT_56_LUNA_2026_0709
             | LMN.AZURE_MODEL_ROUTER_2025_1118
             | LMN.LITELLM_OPENAI_GPT_5
             | LMN.LITELLM_OPENAI_GPT_5_MINI
@@ -215,6 +224,9 @@ def get_encoder_name(model_name: LanguageModelName) -> EncoderName:
             | LMN.LITELLM_OPENAI_GPT_54_THINKING
             | LMN.LITELLM_OPENAI_GPT_55
             | LMN.LITELLM_OPENAI_GPT_55_PRO
+            | LMN.LITELLM_OPENAI_GPT_56_SOL
+            | LMN.LITELLM_OPENAI_GPT_56_TERRA
+            | LMN.LITELLM_OPENAI_GPT_56_LUNA
             | LMN.LITELLM_OPENAI_O1
             | LMN.LITELLM_OPENAI_O3
             | LMN.LITELLM_OPENAI_O3_DEEP_RESEARCH
@@ -1122,6 +1134,46 @@ class LanguageModelInfo(BaseModel):
                     },
                     supported_reasoning_efforts=["medium", "high", "xhigh"],
                 )
+            case (
+                LanguageModelName.AZURE_GPT_56_SOL_2026_0709
+                | LanguageModelName.AZURE_GPT_56_TERRA_2026_0709
+                | LanguageModelName.AZURE_GPT_56_LUNA_2026_0709
+            ):
+                return cls(
+                    name=model_name,
+                    provider=LanguageModelProvider.AZURE,
+                    family=ModelFamily.OPENAI,
+                    version="2026-07-09",
+                    encoder_name=EncoderName.O200K_BASE,
+                    capabilities=[
+                        ModelCapabilities.CHAT_COMPLETIONS_API,
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.PARALLEL_FUNCTION_CALLING,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.RESPONSES_API,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.VISION,
+                    ],
+                    token_limits=LanguageModelTokenLimits(
+                        token_limit_input=922_000, token_limit_output=128_000
+                    ),
+                    info_cutoff_at=date(2026, 6, 1),
+                    published_at=date(2026, 7, 9),
+                    temperature_bounds=TemperatureBounds(
+                        min_temperature=0.0, max_temperature=1.0
+                    ),
+                    default_options={
+                        "reasoning_effort": "none",
+                    },
+                    supported_reasoning_efforts=[
+                        "none",
+                        "low",
+                        "medium",
+                        "high",
+                        "xhigh",
+                    ],
+                )
             case LanguageModelName.AZURE_GPT_4_TURBO_2024_0409:
                 return cls(
                     name=model_name,
@@ -2462,6 +2514,46 @@ class LanguageModelInfo(BaseModel):
                     },
                     supported_reasoning_efforts=["medium", "high", "xhigh"],
                 )
+            case (
+                LanguageModelName.LITELLM_OPENAI_GPT_56_SOL
+                | LanguageModelName.LITELLM_OPENAI_GPT_56_TERRA
+                | LanguageModelName.LITELLM_OPENAI_GPT_56_LUNA
+            ):
+                return cls(
+                    name=model_name,
+                    provider=LanguageModelProvider.LITELLM,
+                    family=ModelFamily.OPENAI,
+                    version="2026-07-09",
+                    encoder_name=EncoderName.O200K_BASE,
+                    capabilities=[
+                        ModelCapabilities.CHAT_COMPLETIONS_API,
+                        ModelCapabilities.FUNCTION_CALLING,
+                        ModelCapabilities.PARALLEL_FUNCTION_CALLING,
+                        ModelCapabilities.REASONING,
+                        ModelCapabilities.RESPONSES_API,
+                        ModelCapabilities.STREAMING,
+                        ModelCapabilities.STRUCTURED_OUTPUT,
+                        ModelCapabilities.VISION,
+                    ],
+                    token_limits=LanguageModelTokenLimits(
+                        token_limit_input=922_000, token_limit_output=128_000
+                    ),
+                    info_cutoff_at=date(2026, 6, 1),
+                    published_at=date(2026, 7, 9),
+                    temperature_bounds=TemperatureBounds(
+                        min_temperature=0.0, max_temperature=1.0
+                    ),
+                    default_options={
+                        "reasoning_effort": "none",
+                    },
+                    supported_reasoning_efforts=[
+                        "none",
+                        "low",
+                        "medium",
+                        "high",
+                        "xhigh",
+                    ],
+                )
             case LanguageModelName.LITELLM_OPENAI_O1:
                 return cls(
                     name=model_name,

{unique_toolkit-2026.30.0.dev10 → unique_toolkit-2026.30.0.dev12}/unique_toolkit/services/chat_service.py

@@ -392,6 +392,7 @@ class ChatService(ChatServiceDeprecated):
         debug_info: dict[str, Any] | None = None,
         message_id: str | None = None,
         set_completed_at: bool | None = None,
+        set_stopped_streaming_at: bool | None = None,
         segment_kind: str | None = None,
     ) -> ChatMessage:
         """Modifies a message in the chat session synchronously if parameter is not specified the corresponding field will remain as is.
@@ -403,6 +404,7 @@ class ChatService(ChatServiceDeprecated):
             debug_info (dict[str, Any]]]): Debug information. Defaults to {}.
             message_id (Optional[str]): The message ID. Defaults to None.
             set_completed_at (Optional[bool]): Whether to set the completedAt field with the current date time. Defaults to False.
+            set_stopped_streaming_at (Optional[bool]): Whether to set the stoppedStreamingAt field with the current date time. Use this to unblock user input as soon as model streaming ends, ahead of slower post-processing (e.g. evaluations) that still needs to run before completedAt is set. Defaults to False.
             segment_kind (str, optional): The segment kind (e.g. "PROCESS", "PREFACE", "ELICITATION", "ANSWER") to relabel the segment to in place, e.g. relabeling M0 from PROCESS to PREFACE once content is set. Defaults to None (kind unchanged).
 
         Returns:
@@ -427,6 +429,7 @@ class ChatService(ChatServiceDeprecated):
             debug_info=debug_info,
             message_id=message_id,
             set_completed_at=set_completed_at or False,
+            set_stopped_streaming_at=set_stopped_streaming_at or False,
             segment_kind=segment_kind,
         )
 
@@ -438,6 +441,7 @@ class ChatService(ChatServiceDeprecated):
         debug_info: dict[str, Any] | None = None,
         message_id: str | None = None,
         set_completed_at: bool | None = False,
+        set_stopped_streaming_at: bool | None = False,
         segment_kind: str | None = None,
     ) -> ChatMessage:
         """Modifies a message in the chat session asynchronously.
@@ -449,6 +453,7 @@ class ChatService(ChatServiceDeprecated):
             references (list[ContentReference]): list of ContentReference objects. Defaults to None.
             debug_info (dict[str, Any]], optional): Debug information. Defaults to None.
             set_completed_at (bool, optional): Whether to set the completedAt field with the current date time. Defaults to False.
+            set_stopped_streaming_at (bool, optional): Whether to set the stoppedStreamingAt field with the current date time. Use this to unblock user input as soon as model streaming ends, ahead of slower post-processing (e.g. evaluations) that still needs to run before completedAt is set. Defaults to False.
             segment_kind (str, optional): The segment kind (e.g. "PROCESS", "PREFACE", "ELICITATION", "ANSWER") to relabel the segment to in place, e.g. relabeling M0 from PROCESS to PREFACE once content is set. Defaults to None (kind unchanged).
 
         Returns:
@@ -472,6 +477,7 @@ class ChatService(ChatServiceDeprecated):
             debug_info=debug_info,
             message_id=message_id,
             set_completed_at=set_completed_at or False,
+            set_stopped_streaming_at=set_stopped_streaming_at or False,
             segment_kind=segment_kind,
         )
 
@@ -655,6 +661,17 @@ class ChatService(ChatServiceDeprecated):
         """Unblocks the next user input"""
         self.modify_assistant_message(set_completed_at=True)
 
+    def signal_streaming_stopped(self) -> None:
+        """Marks model streaming as finished without marking the message fully completed.
+
+        Call this as soon as the model stops emitting tokens, before running any
+        slower post-processing (e.g. evaluations, hallucination checks). Clients
+        that unblock user input on ``stoppedStreamingAt`` (in addition to
+        ``completedAt``) will re-enable input immediately instead of waiting for
+        post-processing to finish.
+        """
+        self.modify_assistant_message(set_stopped_streaming_at=True)
+
     # History Methods
     ############################################################################