claude thinking model support

fix: AWSTranscribeSTTService always set to us-east-1

CHANGELOG.md

@@ -5,10 +5,33 @@ All notable changes to **Pipecat** will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [Unreleased]
+
+### Changed
+
+- Updated `AICFilter` to use Quail STT as the default model
+  (`AICModelType.QUAIL_STT`). Quail STT is optimized for human-to-machine
+  interaction (e.g., voice agents, speech-to-text) and operates at a native
+  sample rate of 16 kHz with fixed enhancement parameters.
+
+### Deprecated
+
+- The `noise_gate_enable` parameter in `AICFilter` is deprecated and no longer
+  has any effect. Noise gating is now handled automatically by the AIC VAD
+  system. Use `AICFilter.create_vad_analyzer()` for VAD functionality instead.
+
+### Fixed
+
+- Fixed an issue in `AWSTranscribeSTTService` where the `region` arg was
+  always set to `us-east-1` when providing an AWS_REGION env var.
+
 ## [0.0.96] - 2025-11-26 🦃 "Happy Thanksgiving!" 🦃
 
 ### Added
 
+- Added `AWSBedrockAgentCoreProcessor` to support invoking an AgentCore-hosted
+  agent in a Pipecat pipeline.
+
 - Enhanced error handling across the framework:
 
   - Added `on_error` callback to `FrameProcessor` for centralized error
@@ -280,7 +303,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Fixed
 
 - Fixed an issue in `AWSBedrockLLMService` where the `aws_region` arg was
-  always set to `us-east-1`.
+  always set to `us-east-1` when providing an AWS_REGION env var.
 
 - Fixed an issue with `DeepgramFluxSTTService` where it sometimes failed to reconnect.
 

examples/foundational/14a-function-calling-anthropic.py

@@ -82,6 +82,13 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
     llm = AnthropicLLMService(
         api_key=os.getenv("ANTHROPIC_API_KEY"),
         model="claude-3-7-sonnet-latest",
+        wait_for_all=True,
+        params=AnthropicLLMService.InputParams(
+            max_tokens=16000,
+            extra={
+                "thinking": {"type": "enabled", "budget_tokens": 10000},
+            },
+        ),
     )
     llm.register_function("get_weather", get_weather)
     llm.register_function("get_restaurant_recommendation", fetch_restaurant_recommendation)

pyproject.toml

@@ -45,7 +45,7 @@ Source = "https://github.com/pipecat-ai/pipecat"
 Website = "https://pipecat.ai"
 
 [project.optional-dependencies]
-aic = [ "aic-sdk~=1.1.0" ]
+aic = [ "aic-sdk~=1.2.0" ]
 anthropic = [ "anthropic~=0.49.0" ]
 assemblyai = [ "pipecat-ai[websockets-base]" ]
 asyncai = [ "pipecat-ai[websockets-base]" ]

src/pipecat/audio/filters/aic_filter.py

@@ -39,7 +39,7 @@ class AICFilter(BaseAudioFilter):
         self,
         *,
         license_key: str = "",
-        model_type: AICModelType = AICModelType.QUAIL_L,
+        model_type: AICModelType = AICModelType.QUAIL_STT,
         enhancement_level: Optional[float] = 1.0,
         voice_gain: Optional[float] = 1.0,
         noise_gate_enable: Optional[bool] = True,
@@ -52,12 +52,27 @@ class AICFilter(BaseAudioFilter):
             enhancement_level: Optional overall enhancement strength (0.0..1.0).
             voice_gain: Optional linear gain applied to detected speech (0.0..4.0).
             noise_gate_enable: Optional enable/disable noise gate (default: True).
+
+                .. deprecated:: 1.3.0
+                    The `noise_gate_enable` parameter is deprecated and no longer has any effect.
+                    It will be removed in a future version.
         """
         self._license_key = license_key
         self._model_type = model_type
 
         self._enhancement_level = enhancement_level
         self._voice_gain = voice_gain
+        if noise_gate_enable is not None:
+            import warnings
+
+            with warnings.catch_warnings():
+                warnings.simplefilter("always")
+                warnings.warn(
+                    "Parameter `noise_gate_enable` is deprecated and no longer has any effect. "
+                    "It will be removed in a future version. Use AIC VAD instead (create_vad_analyzer()).",
+                    DeprecationWarning,
+                )
+
         self._noise_gate_enable = noise_gate_enable
 
         self._enabled = True
@@ -149,10 +164,6 @@ class AICFilter(BaseAudioFilter):
                 )
             if self._voice_gain is not None:
                 self._aic.set_parameter(AICParameter.VOICE_GAIN, float(self._voice_gain))
-            if self._noise_gate_enable is not None:
-                self._aic.set_parameter(
-                    AICParameter.NOISE_GATE_ENABLE, 1.0 if bool(self._noise_gate_enable) else 0.0
-                )
 
             self._aic_ready = True
 

src/pipecat/frames/frames.py

@@ -563,6 +563,33 @@ class LLMContextFrame(Frame):
     context: "LLMContext"
 
 
+@dataclass
+class LLMThinkingTextFrame(DataFrame):
+    """Reasoning frame generated by LLM services."""
+
+    thinking: str
+
+    def __post_init__(self):
+        super().__post_init__()
+        # LLM services send text frames with all necessary spaces included
+        self.includes_inter_frame_spaces = True
+
+    def __str__(self):
+        pts = format_pts(self.pts)
+        return f"{self.name}(pts: {pts}, thinking: {self.thinking})"
+
+
+@dataclass
+class LLMThinkingSignatureFrame(DataFrame):
+    """Reasoning signature frame generated by LLM services."""
+
+    signature: str
+
+    def __str__(self):
+        pts = format_pts(self.pts)
+        return f"{self.name}(pts: {pts}, signature: {self.signature})"
+
+
 @dataclass
 class LLMMessagesFrame(DataFrame):
     """Frame containing LLM messages for chat completion.

src/pipecat/processors/aggregators/llm_response_universal.py

@@ -47,6 +47,8 @@ from pipecat.frames.frames import (
     LLMRunFrame,
     LLMSetToolChoiceFrame,
     LLMSetToolsFrame,
+    LLMThinkingSignatureFrame,
+    LLMThinkingTextFrame,
     SpeechControlParamsFrame,
     StartFrame,
     TextFrame,
@@ -591,6 +593,7 @@ class LLMAssistantAggregator(LLMContextAggregator):
         self._started = 0
         self._function_calls_in_progress: Dict[str, Optional[FunctionCallInProgressFrame]] = {}
         self._context_updated_tasks: Set[asyncio.Task] = set()
+        self._thinking: List[TextPartForConcatenation] = []
 
     @property
     def has_function_calls_in_progress(self) -> bool:
@@ -601,6 +604,11 @@ class LLMAssistantAggregator(LLMContextAggregator):
         """
         return bool(self._function_calls_in_progress)
 
+    async def reset(self):
+        """Reset the aggregation state."""
+        await super().reset()
+        self._thinking = []
+
     async def process_frame(self, frame: Frame, direction: FrameDirection):
         """Process frames for assistant response aggregation and function call management.
 
@@ -619,6 +627,10 @@ class LLMAssistantAggregator(LLMContextAggregator):
             await self._handle_llm_end(frame)
         elif isinstance(frame, TextFrame):
             await self._handle_text(frame)
+        elif isinstance(frame, LLMThinkingTextFrame):
+            await self._handle_thinking(frame)
+        elif isinstance(frame, LLMThinkingSignatureFrame):
+            await self._handle_thinking_signature(frame)
         elif isinstance(frame, LLMRunFrame):
             await self._handle_llm_run(frame)
         elif isinstance(frame, LLMMessagesAppendFrame):
@@ -663,6 +675,14 @@ class LLMAssistantAggregator(LLMContextAggregator):
         timestamp_frame = LLMContextAssistantTimestampFrame(timestamp=time_now_iso8601())
         await self.push_frame(timestamp_frame)
 
+    def thinking_string(self) -> str:
+        """Get the current thinking as a string.
+
+        Returns:
+            The concatenated thinking string.
+        """
+        return concatenate_aggregated_text(self._thinking)
+
     async def _handle_llm_run(self, frame: LLMRunFrame):
         await self.push_context_frame(FrameDirection.UPSTREAM)
 
@@ -824,6 +844,35 @@ class LLMAssistantAggregator(LLMContextAggregator):
             )
         )
 
+    async def _handle_thinking(self, frame: LLMThinkingTextFrame):
+        if not self._started:
+            return
+
+        # Make sure we really have text (spaces count, too!)
+        if len(frame.thinking) == 0:
+            return
+
+        self._thinking.append(
+            TextPartForConcatenation(
+                frame.thinking, includes_inter_part_spaces=frame.includes_inter_frame_spaces
+            )
+        )
+
+    async def _handle_thinking_signature(self, frame: LLMThinkingSignatureFrame):
+        if not self._started:
+            return
+
+        thinking = self.thinking_string()
+
+        self._context.add_message(
+            {
+                "role": "assistant",
+                "content": [
+                    {"type": "thinking", "thinking": thinking, "signature": frame.signature},
+                ],
+            }
+        )
+
     def _context_updated_task_finished(self, task: asyncio.Task):
         self._context_updated_tasks.discard(task)
 

src/pipecat/runner/run.py

@@ -302,7 +302,7 @@ def _setup_webrtc_routes(
         result: StartBotResult = {"sessionId": session_id}
         if request_data.get("enableDefaultIceServers"):
             result["iceConfig"] = IceConfig(
-                iceServers=[IceServer(urls="stun:stun.l.google.com:19302")]
+                iceServers=[IceServer(urls=["stun:stun.l.google.com:19302"])]
             )
 
         return result

src/pipecat/services/anthropic/llm.py

@@ -40,6 +40,8 @@ from pipecat.frames.frames import (
     LLMFullResponseStartFrame,
     LLMMessagesFrame,
     LLMTextFrame,
+    LLMThinkingSignatureFrame,
+    LLMThinkingTextFrame,
     LLMUpdateSettingsFrame,
     UserImageRawFrame,
 )
@@ -380,6 +382,10 @@ class AnthropicLLMService(LLMService):
                         completion_tokens_estimate += self._estimate_tokens(
                             event.delta.partial_json
                         )
+                    elif hasattr(event.delta, "thinking"):
+                        await self.push_frame(LLMThinkingTextFrame(event.delta.thinking))
+                    elif hasattr(event.delta, "signature"):
+                        await self.push_frame(LLMThinkingSignatureFrame(event.delta.signature))
                 elif event.type == "content_block_start":
                     if event.content_block.type == "tool_use":
                         tool_use_block = event.content_block

src/pipecat/services/aws/stt.py

@@ -58,7 +58,7 @@ class AWSTranscribeSTTService(STTService):
         api_key: Optional[str] = None,
         aws_access_key_id: Optional[str] = None,
         aws_session_token: Optional[str] = None,
-        region: Optional[str] = "us-east-1",
+        region: Optional[str] = None,
         sample_rate: int = 16000,
         language: Language = Language.EN,
         **kwargs,
@@ -69,7 +69,7 @@ class AWSTranscribeSTTService(STTService):
             api_key: AWS secret access key. If None, uses AWS_SECRET_ACCESS_KEY environment variable.
             aws_access_key_id: AWS access key ID. If None, uses AWS_ACCESS_KEY_ID environment variable.
             aws_session_token: AWS session token for temporary credentials. If None, uses AWS_SESSION_TOKEN environment variable.
-            region: AWS region for the service. Defaults to "us-east-1".
+            region: AWS region for the service.
             sample_rate: Audio sample rate in Hz. Must be 8000 or 16000. Defaults to 16000.
             language: Language for transcription. Defaults to English.
             **kwargs: Additional arguments passed to parent STTService class.