release: 4.32.0

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "4.31.0"
+  ".": "4.32.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 39
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/retell%2Ftoddlzt-33c92a2f7908f82088c3cad1076ac008f22cc41d25cbdbcd23b005bfd5e5ab3d.yml
-openapi_spec_hash: 3df26020335431e67fdaa9c3865a6a81
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/retell%2Ftoddlzt-01ccd56ecde0b5e2e742326f26f3cf2c0bbafef856e1b032110be4880e162106.yml
+openapi_spec_hash: 8b1688d1e66812ccafbe5fcf26faa67a
 config_hash: f4bc63f2350a2a4988750b41a0737f9d

CHANGELOG.md

@@ -1,5 +1,14 @@
 # Changelog
 
+## 4.32.0 (2025-05-26)
+
+Full Changelog: [v4.31.0...v4.32.0](https://github.com/RetellAI/retell-python-sdk/compare/v4.31.0...v4.32.0)
+
+### Features
+
+* **api:** api update ([59e4d5a](https://github.com/RetellAI/retell-python-sdk/commit/59e4d5a864c6e028437f5f15619aabad7d18b445))
+* **api:** api update ([887256f](https://github.com/RetellAI/retell-python-sdk/commit/887256f2b38d8f6ed6a77d9855a39bc9ee53558f))
+
 ## 4.31.0 (2025-05-22)
 
 Full Changelog: [v4.30.0...v4.31.0](https://github.com/RetellAI/retell-python-sdk/compare/v4.30.0...v4.31.0)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "retell-sdk"
-version = "4.31.0"
+version = "4.32.0"
 description = "The official Python library for the retell API"
 dynamic = ["readme"]
 license = "Apache-2.0"

src/retell/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "retell"
-__version__ = "4.31.0"  # x-release-please-version
+__version__ = "4.32.0"  # x-release-please-version

src/retell/resources/agent.py

@@ -124,6 +124,7 @@ def create(
         stt_mode: Literal["fast", "accurate"] | NotGiven = NOT_GIVEN,
         user_dtmf_options: Optional[agent_create_params.UserDtmfOptions] | NotGiven = NOT_GIVEN,
         version: Optional[float] | NotGiven = NOT_GIVEN,
+        vocab_specialization: Literal["general", "medical"] | NotGiven = NOT_GIVEN,
         voice_model: Optional[
             Literal[
                 "eleven_turbo_v2",
@@ -301,6 +302,10 @@ def create(
 
           version: Version of the agent.
 
+          vocab_specialization: If set, determines the vocabulary set to use for transcription. This setting
+              only applies for English agents, for non English agent, this setting is a no-op.
+              Default to general.
+
           voice_model: Optionally set the voice model used for the selected voice. Currently only
               elevenlab voices have voice model selections. Set to null to remove voice model
               selection, and default ones will apply. Check out the dashboard for details on
@@ -377,6 +382,7 @@ def create(
                     "stt_mode": stt_mode,
                     "user_dtmf_options": user_dtmf_options,
                     "version": version,
+                    "vocab_specialization": vocab_specialization,
                     "voice_model": voice_model,
                     "voice_speed": voice_speed,
                     "voice_temperature": voice_temperature,
@@ -512,6 +518,7 @@ def update(
         stt_mode: Literal["fast", "accurate"] | NotGiven = NOT_GIVEN,
         user_dtmf_options: Optional[agent_update_params.UserDtmfOptions] | NotGiven = NOT_GIVEN,
         body_version: Optional[float] | NotGiven = NOT_GIVEN,
+        vocab_specialization: Literal["general", "medical"] | NotGiven = NOT_GIVEN,
         voice_id: str | NotGiven = NOT_GIVEN,
         voice_model: Optional[
             Literal[
@@ -689,6 +696,10 @@ def update(
 
           body_version: Version of the agent.
 
+          vocab_specialization: If set, determines the vocabulary set to use for transcription. This setting
+              only applies for English agents, for non English agent, this setting is a no-op.
+              Default to general.
+
           voice_id: Unique voice id used for the agent. Find list of available voices and their
               preview in Dashboard.
 
@@ -769,6 +780,7 @@ def update(
                     "stt_mode": stt_mode,
                     "user_dtmf_options": user_dtmf_options,
                     "body_version": body_version,
+                    "vocab_specialization": vocab_specialization,
                     "voice_id": voice_id,
                     "voice_model": voice_model,
                     "voice_speed": voice_speed,
@@ -975,6 +987,7 @@ async def create(
         stt_mode: Literal["fast", "accurate"] | NotGiven = NOT_GIVEN,
         user_dtmf_options: Optional[agent_create_params.UserDtmfOptions] | NotGiven = NOT_GIVEN,
         version: Optional[float] | NotGiven = NOT_GIVEN,
+        vocab_specialization: Literal["general", "medical"] | NotGiven = NOT_GIVEN,
         voice_model: Optional[
             Literal[
                 "eleven_turbo_v2",
@@ -1152,6 +1165,10 @@ async def create(
 
           version: Version of the agent.
 
+          vocab_specialization: If set, determines the vocabulary set to use for transcription. This setting
+              only applies for English agents, for non English agent, this setting is a no-op.
+              Default to general.
+
           voice_model: Optionally set the voice model used for the selected voice. Currently only
               elevenlab voices have voice model selections. Set to null to remove voice model
               selection, and default ones will apply. Check out the dashboard for details on
@@ -1228,6 +1245,7 @@ async def create(
                     "stt_mode": stt_mode,
                     "user_dtmf_options": user_dtmf_options,
                     "version": version,
+                    "vocab_specialization": vocab_specialization,
                     "voice_model": voice_model,
                     "voice_speed": voice_speed,
                     "voice_temperature": voice_temperature,
@@ -1363,6 +1381,7 @@ async def update(
         stt_mode: Literal["fast", "accurate"] | NotGiven = NOT_GIVEN,
         user_dtmf_options: Optional[agent_update_params.UserDtmfOptions] | NotGiven = NOT_GIVEN,
         body_version: Optional[float] | NotGiven = NOT_GIVEN,
+        vocab_specialization: Literal["general", "medical"] | NotGiven = NOT_GIVEN,
         voice_id: str | NotGiven = NOT_GIVEN,
         voice_model: Optional[
             Literal[
@@ -1540,6 +1559,10 @@ async def update(
 
           body_version: Version of the agent.
 
+          vocab_specialization: If set, determines the vocabulary set to use for transcription. This setting
+              only applies for English agents, for non English agent, this setting is a no-op.
+              Default to general.
+
           voice_id: Unique voice id used for the agent. Find list of available voices and their
               preview in Dashboard.
 
@@ -1620,6 +1643,7 @@ async def update(
                     "stt_mode": stt_mode,
                     "user_dtmf_options": user_dtmf_options,
                     "body_version": body_version,
+                    "vocab_specialization": vocab_specialization,
                     "voice_id": voice_id,
                     "voice_model": voice_model,
                     "voice_speed": voice_speed,

src/retell/types/agent_create_params.py

@@ -298,6 +298,13 @@ class AgentCreateParams(TypedDict, total=False):
     version: Optional[float]
     """Version of the agent."""
 
+    vocab_specialization: Literal["general", "medical"]
+    """If set, determines the vocabulary set to use for transcription.
+
+    This setting only applies for English agents, for non English agent, this
+    setting is a no-op. Default to general.
+    """
+
     voice_model: Optional[
         Literal[
             "eleven_turbo_v2",

src/retell/types/agent_response.py

@@ -440,6 +440,13 @@ class AgentResponse(BaseModel):
     version: Optional[object] = None
     """Version of the agent."""
 
+    vocab_specialization: Optional[Literal["general", "medical"]] = None
+    """If set, determines the vocabulary set to use for transcription.
+
+    This setting only applies for English agents, for non English agent, this
+    setting is a no-op. Default to general.
+    """
+
     voice_model: Optional[
         Literal[
             "eleven_turbo_v2",

src/retell/types/agent_update_params.py

@@ -297,6 +297,13 @@ class AgentUpdateParams(TypedDict, total=False):
     body_version: Annotated[Optional[float], PropertyInfo(alias="version")]
     """Version of the agent."""
 
+    vocab_specialization: Literal["general", "medical"]
+    """If set, determines the vocabulary set to use for transcription.
+
+    This setting only applies for English agents, for non English agent, this
+    setting is a no-op. Default to general.
+    """
+
     voice_id: str
     """Unique voice id used for the agent.
 

src/retell/types/chat_response.py

@@ -187,6 +187,9 @@ class ChatResponse(BaseModel):
 
     chat_cost: Optional[ChatCost] = None
 
+    collected_dynamic_variables: Optional[Dict[str, object]] = None
+    """Dynamic variables collected from the chat. Only available after the chat ends."""
+
     end_timestamp: Optional[int] = None
     """End timestamp (milliseconds since epoch) of the chat.
 

src/retell/types/phone_call_response.py

@@ -19,6 +19,7 @@
     "LatencyLlmWebsocketNetworkRtt",
     "LatencyS2s",
     "LatencyTts",
+    "LlmTokenUsage",
     "TelephonyIdentifier",
     "TranscriptObject",
     "TranscriptObjectWord",
@@ -281,6 +282,17 @@ class Latency(BaseModel):
     """
 
 
+class LlmTokenUsage(BaseModel):
+    average: float
+    """Average token count of the call."""
+
+    num_requests: float
+    """Number of requests made to the LLM."""
+
+    values: List[float]
+    """All the token count values in the call."""
+
+
 class TelephonyIdentifier(BaseModel):
     twilio_call_sid: Optional[str] = None
     """Twilio call sid."""
@@ -439,6 +451,9 @@ class PhoneCallResponse(BaseModel):
     call_cost: Optional[CallCost] = None
     """Cost of the call, including all the products and their costs and discount."""
 
+    collected_dynamic_variables: Optional[Dict[str, object]] = None
+    """Dynamic variables collected from the call. Only available after the call ends."""
+
     disconnection_reason: Optional[
         Literal[
             "user_hangup",
@@ -489,6 +504,12 @@ class PhoneCallResponse(BaseModel):
     feature used.
     """
 
+    llm_token_usage: Optional[LlmTokenUsage] = None
+    """LLM token usage of the call, available after call ends.
+
+    Not populated if using custom LLM, realtime API, or no LLM call is made.
+    """
+
     metadata: Optional[object] = None
     """An arbitrary object for storage purpose only.
 

src/retell/types/web_call_response.py

@@ -19,6 +19,7 @@
     "LatencyLlmWebsocketNetworkRtt",
     "LatencyS2s",
     "LatencyTts",
+    "LlmTokenUsage",
     "TranscriptObject",
     "TranscriptObjectWord",
     "TranscriptWithToolCall",
@@ -280,6 +281,17 @@ class Latency(BaseModel):
     """
 
 
+class LlmTokenUsage(BaseModel):
+    average: float
+    """Average token count of the call."""
+
+    num_requests: float
+    """Number of requests made to the LLM."""
+
+    values: List[float]
+    """All the token count values in the call."""
+
+
 class TranscriptObjectWord(BaseModel):
     end: Optional[float] = None
     """End time of the word in the call in second.
@@ -430,6 +442,9 @@ class WebCallResponse(BaseModel):
     call_cost: Optional[CallCost] = None
     """Cost of the call, including all the products and their costs and discount."""
 
+    collected_dynamic_variables: Optional[Dict[str, object]] = None
+    """Dynamic variables collected from the call. Only available after the call ends."""
+
     disconnection_reason: Optional[
         Literal[
             "user_hangup",
@@ -480,6 +495,12 @@ class WebCallResponse(BaseModel):
     feature used.
     """
 
+    llm_token_usage: Optional[LlmTokenUsage] = None
+    """LLM token usage of the call, available after call ends.
+
+    Not populated if using custom LLM, realtime API, or no LLM call is made.
+    """
+
     metadata: Optional[object] = None
     """An arbitrary object for storage purpose only.
 

tests/api_resources/test_agent.py

@@ -88,6 +88,7 @@ def test_method_create_with_all_params(self, client: Retell) -> None:
                 "timeout_ms": 1000,
             },
             version=0,
+            vocab_specialization="general",
             voice_model="eleven_turbo_v2",
             voice_speed=1,
             voice_temperature=1,
@@ -240,6 +241,7 @@ def test_method_update_with_all_params(self, client: Retell) -> None:
                 "timeout_ms": 1000,
             },
             body_version=0,
+            vocab_specialization="general",
             voice_id="11labs-Adrian",
             voice_model="eleven_turbo_v2",
             voice_speed=1,
@@ -454,6 +456,7 @@ async def test_method_create_with_all_params(self, async_client: AsyncRetell) ->
                 "timeout_ms": 1000,
             },
             version=0,
+            vocab_specialization="general",
             voice_model="eleven_turbo_v2",
             voice_speed=1,
             voice_temperature=1,
@@ -606,6 +609,7 @@ async def test_method_update_with_all_params(self, async_client: AsyncRetell) ->
                 "timeout_ms": 1000,
             },
             body_version=0,
+            vocab_specialization="general",
             voice_id="11labs-Adrian",
             voice_model="eleven_turbo_v2",
             voice_speed=1,