Prefer Service.ThinkingConfig over raw ThinkingConfig class names in Anthropic and Google services and examples

examples/foundational/49a-thinking-anthropic.py

@@ -22,7 +22,7 @@ from pipecat.processors.aggregators.llm_response_universal import (
 )
 from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
-from pipecat.services.anthropic.llm import AnthropicLLMService, AnthropicThinkingConfig
+from pipecat.services.anthropic.llm import AnthropicLLMService
 from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.deepgram.stt import DeepgramSTTService
 from pipecat.transports.base_transport import BaseTransport, TransportParams
@@ -64,7 +64,7 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
     llm = AnthropicLLMService(
         api_key=os.getenv("ANTHROPIC_API_KEY"),
         settings=AnthropicLLMService.Settings(
-            thinking=AnthropicThinkingConfig(
+            thinking=AnthropicLLMService.ThinkingConfig(
                 type="enabled",
                 budget_tokens=2048,
             ),

examples/foundational/49b-thinking-google.py

@@ -24,7 +24,7 @@ from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
 from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.deepgram.stt import DeepgramSTTService
-from pipecat.services.google.llm import GoogleLLMService, GoogleThinkingConfig
+from pipecat.services.google.llm import GoogleLLMService
 from pipecat.transports.base_transport import BaseTransport, TransportParams
 from pipecat.transports.daily.transport import DailyParams
 from pipecat.transports.websocket.fastapi import FastAPIWebsocketParams
@@ -65,7 +65,7 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
         api_key=os.getenv("GOOGLE_API_KEY"),
         # model="gemini-3-pro-preview", # A more powerful reasoning model, but slower
         settings=GoogleLLMService.Settings(
-            thinking=GoogleThinkingConfig(
+            thinking=GoogleLLMService.ThinkingConfig(
                 thinking_budget=-1,  # Dynamic thinking
                 include_thoughts=True,
             ),

examples/foundational/49c-thinking-functions-anthropic.py

@@ -23,7 +23,7 @@ from pipecat.processors.aggregators.llm_response_universal import (
 )
 from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
-from pipecat.services.anthropic.llm import AnthropicLLMService, AnthropicThinkingConfig
+from pipecat.services.anthropic.llm import AnthropicLLMService
 from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.deepgram.stt import DeepgramSTTService
 from pipecat.services.llm_service import FunctionCallParams
@@ -85,7 +85,7 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
     llm = AnthropicLLMService(
         api_key=os.getenv("ANTHROPIC_API_KEY"),
         settings=AnthropicLLMService.Settings(
-            thinking=AnthropicThinkingConfig(
+            thinking=AnthropicLLMService.ThinkingConfig(
                 type="enabled",
                 budget_tokens=2048,
             ),

examples/foundational/49d-thinking-functions-google.py

@@ -25,7 +25,7 @@ from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
 from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.deepgram.stt import DeepgramSTTService
-from pipecat.services.google.llm import GoogleLLMService, GoogleThinkingConfig
+from pipecat.services.google.llm import GoogleLLMService
 from pipecat.services.llm_service import FunctionCallParams
 from pipecat.transports.base_transport import BaseTransport, TransportParams
 from pipecat.transports.daily.transport import DailyParams
@@ -86,7 +86,7 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
         api_key=os.getenv("GOOGLE_API_KEY"),
         # model="gemini-3-pro-preview", # A more powerful reasoning model, but slower
         settings=GoogleLLMService.Settings(
-            thinking=GoogleThinkingConfig(
+            thinking=GoogleLLMService.ThinkingConfig(
                 thinking_budget=-1,  # Dynamic thinking
                 include_thoughts=True,
             ),

src/pipecat/services/anthropic/llm.py

@@ -98,18 +98,20 @@ class AnthropicLLMSettings(LLMSettings):
     """
 
     enable_prompt_caching: bool | _NotGiven = field(default_factory=lambda: _NOT_GIVEN)
-    thinking: AnthropicThinkingConfig | _NotGiven = field(default_factory=lambda: _NOT_GIVEN)
+    thinking: Union["AnthropicLLMService.ThinkingConfig", _NotGiven] = field(
+        default_factory=lambda: _NOT_GIVEN
+    )
 
     @classmethod
     def from_mapping(cls, settings):
         """Convert a plain dict to settings, coercing thinking dicts.
 
         For backward compatibility, a ``thinking`` value that is a plain dict
-        is converted to a :class:`AnthropicThinkingConfig`.
+        is converted to a :class:`AnthropicLLMService.ThinkingConfig`.
         """
         instance = super().from_mapping(settings)
         if is_given(instance.thinking) and isinstance(instance.thinking, dict):
-            instance.thinking = AnthropicThinkingConfig(**instance.thinking)
+            instance.thinking = AnthropicLLMService.ThinkingConfig(**instance.thinking)
         return instance
 
 
@@ -199,7 +201,9 @@ class AnthropicLLMService(LLMService):
         temperature: Optional[float] = Field(default_factory=lambda: NOT_GIVEN, ge=0.0, le=1.0)
         top_k: Optional[int] = Field(default_factory=lambda: NOT_GIVEN, ge=0)
         top_p: Optional[float] = Field(default_factory=lambda: NOT_GIVEN, ge=0.0, le=1.0)
-        thinking: Optional[AnthropicThinkingConfig] = Field(default_factory=lambda: NOT_GIVEN)
+        thinking: Optional["AnthropicLLMService.ThinkingConfig"] = Field(
+            default_factory=lambda: NOT_GIVEN
+        )
         extra: Optional[Dict[str, Any]] = Field(default_factory=dict)
 
         def model_post_init(self, __context):

src/pipecat/services/google/llm.py

@@ -16,7 +16,7 @@ import json
 import os
 import uuid
 from dataclasses import dataclass, field
-from typing import Any, AsyncIterator, Dict, List, Literal, Optional
+from typing import Any, AsyncIterator, Dict, List, Literal, Optional, Union
 
 from loguru import logger
 from PIL import Image
@@ -719,18 +719,20 @@ class GoogleLLMSettings(LLMSettings):
         thinking: Thinking configuration.
     """
 
-    thinking: GoogleThinkingConfig | _NotGiven = field(default_factory=lambda: NOT_GIVEN)
+    thinking: Union["GoogleLLMService.ThinkingConfig", _NotGiven] = field(
+        default_factory=lambda: NOT_GIVEN
+    )
 
     @classmethod
     def from_mapping(cls, settings):
         """Convert a plain dict to settings, coercing thinking dicts.
 
         For backward compatibility, a ``thinking`` value that is a plain dict
-        is converted to a :class:`GoogleThinkingConfig`.
+        is converted to a :class:`GoogleLLMService.ThinkingConfig`.
         """
         instance = super().from_mapping(settings)
         if is_given(instance.thinking) and isinstance(instance.thinking, dict):
-            instance.thinking = GoogleThinkingConfig(**instance.thinking)
+            instance.thinking = GoogleLLMService.ThinkingConfig(**instance.thinking)
         return instance
 
 
@@ -775,7 +777,7 @@ class GoogleLLMService(LLMService):
         temperature: Optional[float] = Field(default=None, ge=0.0, le=2.0)
         top_k: Optional[int] = Field(default=None, ge=0)
         top_p: Optional[float] = Field(default=None, ge=0.0, le=1.0)
-        thinking: Optional[GoogleThinkingConfig] = Field(default=None)
+        thinking: Optional["GoogleLLMService.ThinkingConfig"] = Field(default=None)
         extra: Optional[Dict[str, Any]] = Field(default_factory=dict)
 
     def __init__(