diff --git a/langfuse/api/__init__.py b/langfuse/api/__init__.py
index 3a98b0646..ebb0bea02 100644
--- a/langfuse/api/__init__.py
+++ b/langfuse/api/__init__.py
@@ -115,9 +115,11 @@
         PricingTierAttributeCondition,
         PricingTierAttributeSource,
         PricingTierCondition,
+        PricingTierConditionInput,
         PricingTierInput,
         PricingTierOperator,
         PricingTierUsageCondition,
+        PricingTierUsageConditionInput,
         Score,
         ScoreConfig,
         ScoreConfigDataType,
@@ -228,7 +230,7 @@
         PatchMediaBody,
     )
     from .metrics import MetricsV2Response
-    from .models import CreateModelRequest, PaginatedModels
+    from .models import CreateModelRequest, ModelTokenizerId, PaginatedModels
     from .observations import ObservationsV2Meta, ObservationsV2Response
     from .opentelemetry import (
         OtelAttribute,
@@ -501,6 +503,7 @@
     "MetricsV2Response": ".metrics",
     "Model": ".commons",
     "ModelPrice": ".commons",
+    "ModelTokenizerId": ".models",
     "ModelUsageUnit": ".commons",
     "NotFoundError": ".commons",
     "NumericScore": ".commons",
@@ -548,9 +551,11 @@
     "PricingTierAttributeCondition": ".commons",
     "PricingTierAttributeSource": ".commons",
     "PricingTierCondition": ".commons",
+    "PricingTierConditionInput": ".commons",
     "PricingTierInput": ".commons",
     "PricingTierOperator": ".commons",
     "PricingTierUsageCondition": ".commons",
+    "PricingTierUsageConditionInput": ".commons",
     "Project": ".projects",
     "ProjectDeletionResponse": ".projects",
     "Projects": ".projects",
@@ -845,6 +850,7 @@ def __dir__():
     "MetricsV2Response",
     "Model",
     "ModelPrice",
+    "ModelTokenizerId",
     "ModelUsageUnit",
     "NotFoundError",
     "NumericScore",
@@ -892,9 +898,11 @@ def __dir__():
     "PricingTierAttributeCondition",
     "PricingTierAttributeSource",
     "PricingTierCondition",
+    "PricingTierConditionInput",
     "PricingTierInput",
     "PricingTierOperator",
     "PricingTierUsageCondition",
+    "PricingTierUsageConditionInput",
     "Project",
     "ProjectDeletionResponse",
     "Projects",
diff --git a/langfuse/api/commons/__init__.py b/langfuse/api/commons/__init__.py
index 89204bfad..fffe7234d 100644
--- a/langfuse/api/commons/__init__.py
+++ b/langfuse/api/commons/__init__.py
@@ -43,9 +43,11 @@
         PricingTierAttributeCondition,
         PricingTierAttributeSource,
         PricingTierCondition,
+        PricingTierConditionInput,
         PricingTierInput,
         PricingTierOperator,
         PricingTierUsageCondition,
+        PricingTierUsageConditionInput,
         Score,
         ScoreConfig,
         ScoreConfigDataType,
@@ -118,9 +120,11 @@
     "PricingTierAttributeCondition": ".types",
     "PricingTierAttributeSource": ".types",
     "PricingTierCondition": ".types",
+    "PricingTierConditionInput": ".types",
     "PricingTierInput": ".types",
     "PricingTierOperator": ".types",
     "PricingTierUsageCondition": ".types",
+    "PricingTierUsageConditionInput": ".types",
     "Score": ".types",
     "ScoreConfig": ".types",
     "ScoreConfigDataType": ".types",
@@ -216,9 +220,11 @@ def __dir__():
     "PricingTierAttributeCondition",
     "PricingTierAttributeSource",
     "PricingTierCondition",
+    "PricingTierConditionInput",
     "PricingTierInput",
     "PricingTierOperator",
     "PricingTierUsageCondition",
+    "PricingTierUsageConditionInput",
     "Score",
     "ScoreConfig",
     "ScoreConfigDataType",
diff --git a/langfuse/api/commons/types/__init__.py b/langfuse/api/commons/types/__init__.py
index 1bf73059d..6d3303261 100644
--- a/langfuse/api/commons/types/__init__.py
+++ b/langfuse/api/commons/types/__init__.py
@@ -42,9 +42,11 @@
     from .pricing_tier_attribute_condition import PricingTierAttributeCondition
     from .pricing_tier_attribute_source import PricingTierAttributeSource
     from .pricing_tier_condition import PricingTierCondition
+    from .pricing_tier_condition_input import PricingTierConditionInput
     from .pricing_tier_input import PricingTierInput
     from .pricing_tier_operator import PricingTierOperator
     from .pricing_tier_usage_condition import PricingTierUsageCondition
+    from .pricing_tier_usage_condition_input import PricingTierUsageConditionInput
     from .score import (
         Score,
         Score_Boolean,
@@ -109,9 +111,11 @@
     "PricingTierAttributeCondition": ".pricing_tier_attribute_condition",
     "PricingTierAttributeSource": ".pricing_tier_attribute_source",
     "PricingTierCondition": ".pricing_tier_condition",
+    "PricingTierConditionInput": ".pricing_tier_condition_input",
     "PricingTierInput": ".pricing_tier_input",
     "PricingTierOperator": ".pricing_tier_operator",
     "PricingTierUsageCondition": ".pricing_tier_usage_condition",
+    "PricingTierUsageConditionInput": ".pricing_tier_usage_condition_input",
     "Score": ".score",
     "ScoreConfig": ".score_config",
     "ScoreConfigDataType": ".score_config_data_type",
@@ -202,9 +206,11 @@ def __dir__():
     "PricingTierAttributeCondition",
     "PricingTierAttributeSource",
     "PricingTierCondition",
+    "PricingTierConditionInput",
     "PricingTierInput",
     "PricingTierOperator",
     "PricingTierUsageCondition",
+    "PricingTierUsageConditionInput",
     "Score",
     "ScoreConfig",
     "ScoreConfigDataType",
diff --git a/langfuse/api/commons/types/pricing_tier_condition_input.py b/langfuse/api/commons/types/pricing_tier_condition_input.py
new file mode 100644
index 000000000..d7250be3a
--- /dev/null
+++ b/langfuse/api/commons/types/pricing_tier_condition_input.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from .pricing_tier_attribute_condition import PricingTierAttributeCondition
+from .pricing_tier_usage_condition_input import PricingTierUsageConditionInput
+
+PricingTierConditionInput = typing.Union[
+    PricingTierUsageConditionInput, PricingTierAttributeCondition
+]
diff --git a/langfuse/api/commons/types/pricing_tier_input.py b/langfuse/api/commons/types/pricing_tier_input.py
index 40d253ece..6963ed367 100644
--- a/langfuse/api/commons/types/pricing_tier_input.py
+++ b/langfuse/api/commons/types/pricing_tier_input.py
@@ -6,7 +6,7 @@
 import typing_extensions
 from ...core.pydantic_utilities import UniversalBaseModel
 from ...core.serialization import FieldMetadata
-from .pricing_tier_condition import PricingTierCondition
+from .pricing_tier_condition_input import PricingTierConditionInput
 
 
 class PricingTierInput(UniversalBaseModel):
@@ -29,9 +29,9 @@ class PricingTierInput(UniversalBaseModel):
     Must be unique within the model. Common patterns: "Standard", "High Volume Tier", "Extended Context"
     """
 
-    is_default: typing_extensions.Annotated[bool, FieldMetadata(alias="isDefault")] = (
-        pydantic.Field()
-    )
+    is_default: typing_extensions.Annotated[
+        typing.Optional[bool], FieldMetadata(alias="isDefault")
+    ] = pydantic.Field(default=None)
     """
     Whether this is the default tier. Exactly one tier per model must be marked as default.
     
@@ -40,7 +40,7 @@ class PricingTierInput(UniversalBaseModel):
     - Must have priority=0
     - Must have empty conditions array (conditions=[])
     
-    The default tier acts as a fallback when no conditional tiers match.
+    The default tier acts as a fallback when no conditional tiers match. Defaults to false when omitted.
     """
 
     priority: int = pydantic.Field()
@@ -51,7 +51,7 @@ class PricingTierInput(UniversalBaseModel):
     Conditional tiers should use priority 1, 2, 3, etc. based on their specificity.
     """
 
-    conditions: typing.List[PricingTierCondition] = pydantic.Field()
+    conditions: typing.List[PricingTierConditionInput] = pydantic.Field()
     """
     Array of conditions that must ALL be met for this tier to match (AND logic).
     
diff --git a/langfuse/api/commons/types/pricing_tier_usage_condition_input.py b/langfuse/api/commons/types/pricing_tier_usage_condition_input.py
new file mode 100644
index 000000000..07ea86569
--- /dev/null
+++ b/langfuse/api/commons/types/pricing_tier_usage_condition_input.py
@@ -0,0 +1,39 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import pydantic
+import typing_extensions
+from ...core.pydantic_utilities import UniversalBaseModel
+from ...core.serialization import FieldMetadata
+from .pricing_tier_operator import PricingTierOperator
+
+
+class PricingTierUsageConditionInput(UniversalBaseModel):
+    """
+    Input condition that sums usage details whose keys match a regex.
+    """
+
+    usage_detail_pattern: typing_extensions.Annotated[
+        str, FieldMetadata(alias="usageDetailPattern")
+    ] = pydantic.Field()
+    """
+    Regex pattern matched against usage detail keys.
+    """
+
+    operator: PricingTierOperator
+    value: float = pydantic.Field()
+    """
+    Numeric threshold for the summed matching usage values.
+    """
+
+    case_sensitive: typing_extensions.Annotated[
+        typing.Optional[bool], FieldMetadata(alias="caseSensitive")
+    ] = pydantic.Field(default=None)
+    """
+    Whether the usage-detail regex is case-sensitive. Defaults to false.
+    """
+
+    model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(
+        extra="allow", frozen=True
+    )
diff --git a/langfuse/api/models/__init__.py b/langfuse/api/models/__init__.py
index 7ebdb7762..a027e8f49 100644
--- a/langfuse/api/models/__init__.py
+++ b/langfuse/api/models/__init__.py
@@ -6,9 +6,10 @@
 from importlib import import_module
 
 if typing.TYPE_CHECKING:
-    from .types import CreateModelRequest, PaginatedModels
+    from .types import CreateModelRequest, ModelTokenizerId, PaginatedModels
 _dynamic_imports: typing.Dict[str, str] = {
     "CreateModelRequest": ".types",
+    "ModelTokenizerId": ".types",
     "PaginatedModels": ".types",
 }
 
@@ -40,4 +41,4 @@ def __dir__():
     return sorted(lazy_attrs)
 
 
-__all__ = ["CreateModelRequest", "PaginatedModels"]
+__all__ = ["CreateModelRequest", "ModelTokenizerId", "PaginatedModels"]
diff --git a/langfuse/api/models/client.py b/langfuse/api/models/client.py
index 9f817b8f6..05b245e5b 100644
--- a/langfuse/api/models/client.py
+++ b/langfuse/api/models/client.py
@@ -9,6 +9,7 @@
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .raw_client import AsyncRawModelsClient, RawModelsClient
+from .types.model_tokenizer_id import ModelTokenizerId
 from .types.paginated_models import PaginatedModels
 
 # this is used as the default value for optional parameters
@@ -35,13 +36,13 @@ def create(
         *,
         model_name: str,
         match_pattern: str,
+        unit: ModelUsageUnit,
         start_date: typing.Optional[dt.datetime] = OMIT,
-        unit: typing.Optional[ModelUsageUnit] = OMIT,
         input_price: typing.Optional[float] = OMIT,
         output_price: typing.Optional[float] = OMIT,
         total_price: typing.Optional[float] = OMIT,
         pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
-        tokenizer_id: typing.Optional[str] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
         tokenizer_config: typing.Optional[typing.Any] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> Model:
@@ -56,12 +57,12 @@ def create(
         match_pattern : str
             Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
 
+        unit : ModelUsageUnit
+            Unit used by this model.
+
         start_date : typing.Optional[dt.datetime]
             Apply only to generations which are newer than this ISO date.
 
-        unit : typing.Optional[ModelUsageUnit]
-            Unit used by this model.
-
         input_price : typing.Optional[float]
             Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
 
@@ -91,7 +92,7 @@ def create(
             If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
             which will automatically create a single default tier named "Standard".
 
-        tokenizer_id : typing.Optional[str]
+        tokenizer_id : typing.Optional[ModelTokenizerId]
             Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
 
         tokenizer_config : typing.Optional[typing.Any]
@@ -107,6 +108,7 @@ def create(
         Examples
         --------
         from langfuse import LangfuseAPI
+        from langfuse.commons import ModelUsageUnit
 
         client = LangfuseAPI(
             x_langfuse_sdk_name="YOUR_X_LANGFUSE_SDK_NAME",
@@ -119,13 +121,127 @@ def create(
         client.models.create(
             model_name="modelName",
             match_pattern="matchPattern",
+            unit=ModelUsageUnit.CHARACTERS,
         )
         """
         _response = self._raw_client.create(
             model_name=model_name,
             match_pattern=match_pattern,
+            unit=unit,
             start_date=start_date,
+            input_price=input_price,
+            output_price=output_price,
+            total_price=total_price,
+            pricing_tiers=pricing_tiers,
+            tokenizer_id=tokenizer_id,
+            tokenizer_config=tokenizer_config,
+            request_options=request_options,
+        )
+        return _response.data
+
+    def upsert(
+        self,
+        id: str,
+        *,
+        model_name: str,
+        match_pattern: str,
+        unit: ModelUsageUnit,
+        start_date: typing.Optional[dt.datetime] = OMIT,
+        input_price: typing.Optional[float] = OMIT,
+        output_price: typing.Optional[float] = OMIT,
+        total_price: typing.Optional[float] = OMIT,
+        pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
+        tokenizer_config: typing.Optional[typing.Any] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> Model:
+        """
+        Create or replace a project-owned model using its id. Built-in models cannot be modified.
+
+        Parameters
+        ----------
+        id : str
+
+        model_name : str
+            Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime
+
+        match_pattern : str
+            Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
+
+        unit : ModelUsageUnit
+            Unit used by this model.
+
+        start_date : typing.Optional[dt.datetime]
+            Apply only to generations which are newer than this ISO date.
+
+        input_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
+
+        output_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per output unit. Creates a default tier if pricingTiers not provided.
+
+        total_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per total units. Cannot be set if input or output price is set. Creates a default tier if pricingTiers not provided.
+
+        pricing_tiers : typing.Optional[typing.Sequence[PricingTierInput]]
+            Optional. Array of pricing tiers for this model.
+
+            Use pricing tiers for all models - both those with threshold-based pricing variations and those with simple flat pricing:
+
+            - For models with standard flat pricing: Create a single default tier with your prices
+              (e.g., one tier with isDefault=true, priority=0, conditions=[], and your standard prices)
+
+            - For models with threshold-based pricing: Create a default tier plus additional conditional tiers
+              (e.g., default tier for standard usage + high-volume tier for usage above certain thresholds)
+
+            Requirements:
+            - Cannot be provided with flat prices (inputPrice/outputPrice/totalPrice) - use one approach or the other
+            - Must include exactly one default tier with isDefault=true, priority=0, and conditions=[]
+            - All tier names and priorities must be unique within the model
+            - Each tier must define at least one price
+
+            If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
+            which will automatically create a single default tier named "Standard".
+
+        tokenizer_id : typing.Optional[ModelTokenizerId]
+            Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+
+        tokenizer_config : typing.Optional[typing.Any]
+            Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        Model
+
+        Examples
+        --------
+        from langfuse import LangfuseAPI
+        from langfuse.commons import ModelUsageUnit
+
+        client = LangfuseAPI(
+            x_langfuse_sdk_name="YOUR_X_LANGFUSE_SDK_NAME",
+            x_langfuse_sdk_version="YOUR_X_LANGFUSE_SDK_VERSION",
+            x_langfuse_public_key="YOUR_X_LANGFUSE_PUBLIC_KEY",
+            username="YOUR_USERNAME",
+            password="YOUR_PASSWORD",
+            base_url="https://yourhost.com/path/to/api",
+        )
+        client.models.upsert(
+            id="id",
+            model_name="modelName",
+            match_pattern="matchPattern",
+            unit=ModelUsageUnit.CHARACTERS,
+        )
+        """
+        _response = self._raw_client.upsert(
+            id,
+            model_name=model_name,
+            match_pattern=match_pattern,
             unit=unit,
+            start_date=start_date,
             input_price=input_price,
             output_price=output_price,
             total_price=total_price,
@@ -273,13 +389,13 @@ async def create(
         *,
         model_name: str,
         match_pattern: str,
+        unit: ModelUsageUnit,
         start_date: typing.Optional[dt.datetime] = OMIT,
-        unit: typing.Optional[ModelUsageUnit] = OMIT,
         input_price: typing.Optional[float] = OMIT,
         output_price: typing.Optional[float] = OMIT,
         total_price: typing.Optional[float] = OMIT,
         pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
-        tokenizer_id: typing.Optional[str] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
         tokenizer_config: typing.Optional[typing.Any] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> Model:
@@ -294,12 +410,12 @@ async def create(
         match_pattern : str
             Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
 
+        unit : ModelUsageUnit
+            Unit used by this model.
+
         start_date : typing.Optional[dt.datetime]
             Apply only to generations which are newer than this ISO date.
 
-        unit : typing.Optional[ModelUsageUnit]
-            Unit used by this model.
-
         input_price : typing.Optional[float]
             Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
 
@@ -329,7 +445,7 @@ async def create(
             If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
             which will automatically create a single default tier named "Standard".
 
-        tokenizer_id : typing.Optional[str]
+        tokenizer_id : typing.Optional[ModelTokenizerId]
             Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
 
         tokenizer_config : typing.Optional[typing.Any]
@@ -347,6 +463,7 @@ async def create(
         import asyncio
 
         from langfuse import AsyncLangfuseAPI
+        from langfuse.commons import ModelUsageUnit
 
         client = AsyncLangfuseAPI(
             x_langfuse_sdk_name="YOUR_X_LANGFUSE_SDK_NAME",
@@ -362,6 +479,7 @@ async def main() -> None:
             await client.models.create(
                 model_name="modelName",
                 match_pattern="matchPattern",
+                unit=ModelUsageUnit.CHARACTERS,
             )
 
 
@@ -370,8 +488,129 @@ async def main() -> None:
         _response = await self._raw_client.create(
             model_name=model_name,
             match_pattern=match_pattern,
+            unit=unit,
             start_date=start_date,
+            input_price=input_price,
+            output_price=output_price,
+            total_price=total_price,
+            pricing_tiers=pricing_tiers,
+            tokenizer_id=tokenizer_id,
+            tokenizer_config=tokenizer_config,
+            request_options=request_options,
+        )
+        return _response.data
+
+    async def upsert(
+        self,
+        id: str,
+        *,
+        model_name: str,
+        match_pattern: str,
+        unit: ModelUsageUnit,
+        start_date: typing.Optional[dt.datetime] = OMIT,
+        input_price: typing.Optional[float] = OMIT,
+        output_price: typing.Optional[float] = OMIT,
+        total_price: typing.Optional[float] = OMIT,
+        pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
+        tokenizer_config: typing.Optional[typing.Any] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> Model:
+        """
+        Create or replace a project-owned model using its id. Built-in models cannot be modified.
+
+        Parameters
+        ----------
+        id : str
+
+        model_name : str
+            Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime
+
+        match_pattern : str
+            Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
+
+        unit : ModelUsageUnit
+            Unit used by this model.
+
+        start_date : typing.Optional[dt.datetime]
+            Apply only to generations which are newer than this ISO date.
+
+        input_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
+
+        output_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per output unit. Creates a default tier if pricingTiers not provided.
+
+        total_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per total units. Cannot be set if input or output price is set. Creates a default tier if pricingTiers not provided.
+
+        pricing_tiers : typing.Optional[typing.Sequence[PricingTierInput]]
+            Optional. Array of pricing tiers for this model.
+
+            Use pricing tiers for all models - both those with threshold-based pricing variations and those with simple flat pricing:
+
+            - For models with standard flat pricing: Create a single default tier with your prices
+              (e.g., one tier with isDefault=true, priority=0, conditions=[], and your standard prices)
+
+            - For models with threshold-based pricing: Create a default tier plus additional conditional tiers
+              (e.g., default tier for standard usage + high-volume tier for usage above certain thresholds)
+
+            Requirements:
+            - Cannot be provided with flat prices (inputPrice/outputPrice/totalPrice) - use one approach or the other
+            - Must include exactly one default tier with isDefault=true, priority=0, and conditions=[]
+            - All tier names and priorities must be unique within the model
+            - Each tier must define at least one price
+
+            If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
+            which will automatically create a single default tier named "Standard".
+
+        tokenizer_id : typing.Optional[ModelTokenizerId]
+            Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+
+        tokenizer_config : typing.Optional[typing.Any]
+            Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        Model
+
+        Examples
+        --------
+        import asyncio
+
+        from langfuse import AsyncLangfuseAPI
+        from langfuse.commons import ModelUsageUnit
+
+        client = AsyncLangfuseAPI(
+            x_langfuse_sdk_name="YOUR_X_LANGFUSE_SDK_NAME",
+            x_langfuse_sdk_version="YOUR_X_LANGFUSE_SDK_VERSION",
+            x_langfuse_public_key="YOUR_X_LANGFUSE_PUBLIC_KEY",
+            username="YOUR_USERNAME",
+            password="YOUR_PASSWORD",
+            base_url="https://yourhost.com/path/to/api",
+        )
+
+
+        async def main() -> None:
+            await client.models.upsert(
+                id="id",
+                model_name="modelName",
+                match_pattern="matchPattern",
+                unit=ModelUsageUnit.CHARACTERS,
+            )
+
+
+        asyncio.run(main())
+        """
+        _response = await self._raw_client.upsert(
+            id,
+            model_name=model_name,
+            match_pattern=match_pattern,
             unit=unit,
+            start_date=start_date,
             input_price=input_price,
             output_price=output_price,
             total_price=total_price,
diff --git a/langfuse/api/models/raw_client.py b/langfuse/api/models/raw_client.py
index 0fdc72319..4d645dcab 100644
--- a/langfuse/api/models/raw_client.py
+++ b/langfuse/api/models/raw_client.py
@@ -19,6 +19,7 @@
 from ..core.pydantic_utilities import parse_obj_as
 from ..core.request_options import RequestOptions
 from ..core.serialization import convert_and_respect_annotation_metadata
+from .types.model_tokenizer_id import ModelTokenizerId
 from .types.paginated_models import PaginatedModels
 
 # this is used as the default value for optional parameters
@@ -34,13 +35,13 @@ def create(
         *,
         model_name: str,
         match_pattern: str,
+        unit: ModelUsageUnit,
         start_date: typing.Optional[dt.datetime] = OMIT,
-        unit: typing.Optional[ModelUsageUnit] = OMIT,
         input_price: typing.Optional[float] = OMIT,
         output_price: typing.Optional[float] = OMIT,
         total_price: typing.Optional[float] = OMIT,
         pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
-        tokenizer_id: typing.Optional[str] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
         tokenizer_config: typing.Optional[typing.Any] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> HttpResponse[Model]:
@@ -55,12 +56,12 @@ def create(
         match_pattern : str
             Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
 
+        unit : ModelUsageUnit
+            Unit used by this model.
+
         start_date : typing.Optional[dt.datetime]
             Apply only to generations which are newer than this ISO date.
 
-        unit : typing.Optional[ModelUsageUnit]
-            Unit used by this model.
-
         input_price : typing.Optional[float]
             Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
 
@@ -90,7 +91,7 @@ def create(
             If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
             which will automatically create a single default tier named "Standard".
 
-        tokenizer_id : typing.Optional[str]
+        tokenizer_id : typing.Optional[ModelTokenizerId]
             Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
 
         tokenizer_config : typing.Optional[typing.Any]
@@ -116,7 +117,184 @@ def create(
                 "totalPrice": total_price,
                 "pricingTiers": convert_and_respect_annotation_metadata(
                     object_=pricing_tiers,
-                    annotation=typing.Sequence[PricingTierInput],
+                    annotation=typing.Optional[typing.Sequence[PricingTierInput]],
+                    direction="write",
+                ),
+                "tokenizerId": tokenizer_id,
+                "tokenizerConfig": tokenizer_config,
+            },
+            request_options=request_options,
+            omit=OMIT,
+        )
+        try:
+            if 200 <= _response.status_code < 300:
+                _data = typing.cast(
+                    Model,
+                    parse_obj_as(
+                        type_=Model,  # type: ignore
+                        object_=_response.json(),
+                    ),
+                )
+                return HttpResponse(response=_response, data=_data)
+            if _response.status_code == 400:
+                raise Error(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 401:
+                raise UnauthorizedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 403:
+                raise AccessDeniedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 405:
+                raise MethodNotAllowedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 404:
+                raise NotFoundError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(
+                status_code=_response.status_code,
+                headers=dict(_response.headers),
+                body=_response.text,
+            )
+        raise ApiError(
+            status_code=_response.status_code,
+            headers=dict(_response.headers),
+            body=_response_json,
+        )
+
+    def upsert(
+        self,
+        id: str,
+        *,
+        model_name: str,
+        match_pattern: str,
+        unit: ModelUsageUnit,
+        start_date: typing.Optional[dt.datetime] = OMIT,
+        input_price: typing.Optional[float] = OMIT,
+        output_price: typing.Optional[float] = OMIT,
+        total_price: typing.Optional[float] = OMIT,
+        pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
+        tokenizer_config: typing.Optional[typing.Any] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> HttpResponse[Model]:
+        """
+        Create or replace a project-owned model using its id. Built-in models cannot be modified.
+
+        Parameters
+        ----------
+        id : str
+
+        model_name : str
+            Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime
+
+        match_pattern : str
+            Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
+
+        unit : ModelUsageUnit
+            Unit used by this model.
+
+        start_date : typing.Optional[dt.datetime]
+            Apply only to generations which are newer than this ISO date.
+
+        input_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
+
+        output_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per output unit. Creates a default tier if pricingTiers not provided.
+
+        total_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per total units. Cannot be set if input or output price is set. Creates a default tier if pricingTiers not provided.
+
+        pricing_tiers : typing.Optional[typing.Sequence[PricingTierInput]]
+            Optional. Array of pricing tiers for this model.
+
+            Use pricing tiers for all models - both those with threshold-based pricing variations and those with simple flat pricing:
+
+            - For models with standard flat pricing: Create a single default tier with your prices
+              (e.g., one tier with isDefault=true, priority=0, conditions=[], and your standard prices)
+
+            - For models with threshold-based pricing: Create a default tier plus additional conditional tiers
+              (e.g., default tier for standard usage + high-volume tier for usage above certain thresholds)
+
+            Requirements:
+            - Cannot be provided with flat prices (inputPrice/outputPrice/totalPrice) - use one approach or the other
+            - Must include exactly one default tier with isDefault=true, priority=0, and conditions=[]
+            - All tier names and priorities must be unique within the model
+            - Each tier must define at least one price
+
+            If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
+            which will automatically create a single default tier named "Standard".
+
+        tokenizer_id : typing.Optional[ModelTokenizerId]
+            Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+
+        tokenizer_config : typing.Optional[typing.Any]
+            Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        HttpResponse[Model]
+        """
+        _response = self._client_wrapper.httpx_client.request(
+            f"api/public/models/{jsonable_encoder(id)}",
+            method="PUT",
+            json={
+                "modelName": model_name,
+                "matchPattern": match_pattern,
+                "startDate": start_date,
+                "unit": unit,
+                "inputPrice": input_price,
+                "outputPrice": output_price,
+                "totalPrice": total_price,
+                "pricingTiers": convert_and_respect_annotation_metadata(
+                    object_=pricing_tiers,
+                    annotation=typing.Optional[typing.Sequence[PricingTierInput]],
                     direction="write",
                 ),
                 "tokenizerId": tokenizer_id,
@@ -518,13 +696,13 @@ async def create(
         *,
         model_name: str,
         match_pattern: str,
+        unit: ModelUsageUnit,
         start_date: typing.Optional[dt.datetime] = OMIT,
-        unit: typing.Optional[ModelUsageUnit] = OMIT,
         input_price: typing.Optional[float] = OMIT,
         output_price: typing.Optional[float] = OMIT,
         total_price: typing.Optional[float] = OMIT,
         pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
-        tokenizer_id: typing.Optional[str] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
         tokenizer_config: typing.Optional[typing.Any] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> AsyncHttpResponse[Model]:
@@ -539,12 +717,12 @@ async def create(
         match_pattern : str
             Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
 
+        unit : ModelUsageUnit
+            Unit used by this model.
+
         start_date : typing.Optional[dt.datetime]
             Apply only to generations which are newer than this ISO date.
 
-        unit : typing.Optional[ModelUsageUnit]
-            Unit used by this model.
-
         input_price : typing.Optional[float]
             Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
 
@@ -574,7 +752,7 @@ async def create(
             If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
             which will automatically create a single default tier named "Standard".
 
-        tokenizer_id : typing.Optional[str]
+        tokenizer_id : typing.Optional[ModelTokenizerId]
             Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
 
         tokenizer_config : typing.Optional[typing.Any]
@@ -600,7 +778,184 @@ async def create(
                 "totalPrice": total_price,
                 "pricingTiers": convert_and_respect_annotation_metadata(
                     object_=pricing_tiers,
-                    annotation=typing.Sequence[PricingTierInput],
+                    annotation=typing.Optional[typing.Sequence[PricingTierInput]],
+                    direction="write",
+                ),
+                "tokenizerId": tokenizer_id,
+                "tokenizerConfig": tokenizer_config,
+            },
+            request_options=request_options,
+            omit=OMIT,
+        )
+        try:
+            if 200 <= _response.status_code < 300:
+                _data = typing.cast(
+                    Model,
+                    parse_obj_as(
+                        type_=Model,  # type: ignore
+                        object_=_response.json(),
+                    ),
+                )
+                return AsyncHttpResponse(response=_response, data=_data)
+            if _response.status_code == 400:
+                raise Error(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 401:
+                raise UnauthorizedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 403:
+                raise AccessDeniedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 405:
+                raise MethodNotAllowedError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            if _response.status_code == 404:
+                raise NotFoundError(
+                    headers=dict(_response.headers),
+                    body=typing.cast(
+                        typing.Any,
+                        parse_obj_as(
+                            type_=typing.Any,  # type: ignore
+                            object_=_response.json(),
+                        ),
+                    ),
+                )
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(
+                status_code=_response.status_code,
+                headers=dict(_response.headers),
+                body=_response.text,
+            )
+        raise ApiError(
+            status_code=_response.status_code,
+            headers=dict(_response.headers),
+            body=_response_json,
+        )
+
+    async def upsert(
+        self,
+        id: str,
+        *,
+        model_name: str,
+        match_pattern: str,
+        unit: ModelUsageUnit,
+        start_date: typing.Optional[dt.datetime] = OMIT,
+        input_price: typing.Optional[float] = OMIT,
+        output_price: typing.Optional[float] = OMIT,
+        total_price: typing.Optional[float] = OMIT,
+        pricing_tiers: typing.Optional[typing.Sequence[PricingTierInput]] = OMIT,
+        tokenizer_id: typing.Optional[ModelTokenizerId] = OMIT,
+        tokenizer_config: typing.Optional[typing.Any] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> AsyncHttpResponse[Model]:
+        """
+        Create or replace a project-owned model using its id. Built-in models cannot be modified.
+
+        Parameters
+        ----------
+        id : str
+
+        model_name : str
+            Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime
+
+        match_pattern : str
+            Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`
+
+        unit : ModelUsageUnit
+            Unit used by this model.
+
+        start_date : typing.Optional[dt.datetime]
+            Apply only to generations which are newer than this ISO date.
+
+        input_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per input unit. Creates a default tier if pricingTiers not provided.
+
+        output_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per output unit. Creates a default tier if pricingTiers not provided.
+
+        total_price : typing.Optional[float]
+            Deprecated. Use 'pricingTiers' instead. Price (USD) per total units. Cannot be set if input or output price is set. Creates a default tier if pricingTiers not provided.
+
+        pricing_tiers : typing.Optional[typing.Sequence[PricingTierInput]]
+            Optional. Array of pricing tiers for this model.
+
+            Use pricing tiers for all models - both those with threshold-based pricing variations and those with simple flat pricing:
+
+            - For models with standard flat pricing: Create a single default tier with your prices
+              (e.g., one tier with isDefault=true, priority=0, conditions=[], and your standard prices)
+
+            - For models with threshold-based pricing: Create a default tier plus additional conditional tiers
+              (e.g., default tier for standard usage + high-volume tier for usage above certain thresholds)
+
+            Requirements:
+            - Cannot be provided with flat prices (inputPrice/outputPrice/totalPrice) - use one approach or the other
+            - Must include exactly one default tier with isDefault=true, priority=0, and conditions=[]
+            - All tier names and priorities must be unique within the model
+            - Each tier must define at least one price
+
+            If omitted, you must provide flat prices instead (inputPrice/outputPrice/totalPrice),
+            which will automatically create a single default tier named "Standard".
+
+        tokenizer_id : typing.Optional[ModelTokenizerId]
+            Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+
+        tokenizer_config : typing.Optional[typing.Any]
+            Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        AsyncHttpResponse[Model]
+        """
+        _response = await self._client_wrapper.httpx_client.request(
+            f"api/public/models/{jsonable_encoder(id)}",
+            method="PUT",
+            json={
+                "modelName": model_name,
+                "matchPattern": match_pattern,
+                "startDate": start_date,
+                "unit": unit,
+                "inputPrice": input_price,
+                "outputPrice": output_price,
+                "totalPrice": total_price,
+                "pricingTiers": convert_and_respect_annotation_metadata(
+                    object_=pricing_tiers,
+                    annotation=typing.Optional[typing.Sequence[PricingTierInput]],
                     direction="write",
                 ),
                 "tokenizerId": tokenizer_id,
diff --git a/langfuse/api/models/types/__init__.py b/langfuse/api/models/types/__init__.py
index 8b4b651c5..80eb9ece1 100644
--- a/langfuse/api/models/types/__init__.py
+++ b/langfuse/api/models/types/__init__.py
@@ -7,9 +7,11 @@
 
 if typing.TYPE_CHECKING:
     from .create_model_request import CreateModelRequest
+    from .model_tokenizer_id import ModelTokenizerId
     from .paginated_models import PaginatedModels
 _dynamic_imports: typing.Dict[str, str] = {
     "CreateModelRequest": ".create_model_request",
+    "ModelTokenizerId": ".model_tokenizer_id",
     "PaginatedModels": ".paginated_models",
 }
 
@@ -41,4 +43,4 @@ def __dir__():
     return sorted(lazy_attrs)
 
 
-__all__ = ["CreateModelRequest", "PaginatedModels"]
+__all__ = ["CreateModelRequest", "ModelTokenizerId", "PaginatedModels"]
diff --git a/langfuse/api/models/types/create_model_request.py b/langfuse/api/models/types/create_model_request.py
index dc19db944..458a5290a 100644
--- a/langfuse/api/models/types/create_model_request.py
+++ b/langfuse/api/models/types/create_model_request.py
@@ -9,6 +9,7 @@
 from ...commons.types.pricing_tier_input import PricingTierInput
 from ...core.pydantic_utilities import UniversalBaseModel
 from ...core.serialization import FieldMetadata
+from .model_tokenizer_id import ModelTokenizerId
 
 
 class CreateModelRequest(UniversalBaseModel):
@@ -33,7 +34,7 @@ class CreateModelRequest(UniversalBaseModel):
     Apply only to generations which are newer than this ISO date.
     """
 
-    unit: typing.Optional[ModelUsageUnit] = pydantic.Field(default=None)
+    unit: ModelUsageUnit = pydantic.Field()
     """
     Unit used by this model.
     """
@@ -85,7 +86,7 @@ class CreateModelRequest(UniversalBaseModel):
     """
 
     tokenizer_id: typing_extensions.Annotated[
-        typing.Optional[str], FieldMetadata(alias="tokenizerId")
+        typing.Optional[ModelTokenizerId], FieldMetadata(alias="tokenizerId")
     ] = pydantic.Field(default=None)
     """
     Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
diff --git a/langfuse/api/models/types/model_tokenizer_id.py b/langfuse/api/models/types/model_tokenizer_id.py
new file mode 100644
index 000000000..83a9b234d
--- /dev/null
+++ b/langfuse/api/models/types/model_tokenizer_id.py
@@ -0,0 +1,26 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ...core import enum
+
+T_Result = typing.TypeVar("T_Result")
+
+
+class ModelTokenizerId(enum.StrEnum):
+    """
+    Tokenizer supported by Langfuse for model usage inference.
+    """
+
+    OPENAI = "openai"
+    CLAUDE = "claude"
+
+    def visit(
+        self,
+        openai: typing.Callable[[], T_Result],
+        claude: typing.Callable[[], T_Result],
+    ) -> T_Result:
+        if self is ModelTokenizerId.OPENAI:
+            return openai()
+        if self is ModelTokenizerId.CLAUDE:
+            return claude()