Merge pull request #1296 from pipecat-ai/aleix/google-always-send-text-with-audio

GoogleLLMService: always send text with audio

CHANGELOG.md

@@ -82,6 +82,9 @@ stt = DeepgramSTTService(..., live_options=LiveOptions(model="nova-2-general"))
 
 ### Fixed
 
+- Fixed a `GoogleLLMService` that was causing an exception when sending inline
+  audio in some cases.
+
 - Fixed an `AudioContextWordTTSService` issue that would cause an `EndFrame` to
   disconnect from the TTS service before audio from all the contexts was
   received. This affected services like Cartesia and Rime.

examples/foundational/22d-natural-conversation-gemini-audio.py

@@ -389,7 +389,7 @@ class AudioAccumulator(FrameProcessor):
             )
             self._user_speaking = False
             context = GoogleLLMContext()
-            context.add_audio_frames_message(text="Audio follows", audio_frames=self._audio_frames)
+            context.add_audio_frames_message(audio_frames=self._audio_frames)
             await self.push_frame(OpenAILLMContextFrame(context=context))
         elif isinstance(frame, InputAudioRawFrame):
             # Append the audio frame to our buffer. Treat the buffer as a ring buffer, dropping the oldest

examples/phone-chatbot/bot_daily_gemini.py

@@ -7,20 +7,15 @@ import argparse
 import asyncio
 import os
 import sys
-from dataclasses import dataclass
 from typing import Optional
 
-import google.ai.generativelanguage as glm
 from dotenv import load_dotenv
 from loguru import logger
 
 from pipecat.audio.vad.silero import SileroVADAnalyzer
 from pipecat.frames.frames import (
-    BotStoppedSpeakingFrame,
     EndTaskFrame,
-    Frame,
     InputAudioRawFrame,
-    SystemFrame,
     TranscriptionFrame,
     UserStartedSpeakingFrame,
     UserStoppedSpeakingFrame,
@@ -28,11 +23,11 @@ from pipecat.frames.frames import (
 from pipecat.pipeline.pipeline import Pipeline
 from pipecat.pipeline.runner import PipelineRunner
 from pipecat.pipeline.task import PipelineParams, PipelineTask
-from pipecat.processors.aggregators.openai_llm_context import OpenAILLMContextFrame
 from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
 from pipecat.services.ai_services import LLMService
 from pipecat.services.elevenlabs import ElevenLabsTTSService
-from pipecat.services.google import GoogleLLMContext, GoogleLLMService
+from pipecat.services.google import GoogleLLMService
+from pipecat.services.google.google import GoogleLLMContext
 from pipecat.transports.services.daily import DailyDialinSettings, DailyParams, DailyTransport
 
 load_dotenv(override=True)
@@ -240,7 +235,7 @@ If it sounds like a human (saying hello, asking questions, etc.), call the funct
 DO NOT say anything until you've determined if this is a voicemail or human."""
 
     llm = GoogleLLMService(
-        model="models/gemini-2.0-flash-lite-preview-02-05",
+        model="models/gemini-2.0-flash-lite",
         api_key=os.getenv("GOOGLE_API_KEY"),
         system_instruction=system_instruction,
         tools=tools,

pyproject.toml

@@ -22,9 +22,6 @@ classifiers = [
 dependencies = [
     "aiohttp~=3.11.11",
     "audioop-lts~=0.2.1; python_version>='3.13'",
-    # We need an older version of `httpx` that doesn't remove the deprecated
-    # `proxies` argument. This is necessary for Azure and Anthropic clients.
-    "httpx~=0.27.2",
     "loguru~=0.7.3",
     "Markdown~=3.7",
     "numpy~=1.26.4",
@@ -42,7 +39,7 @@ Source = "https://github.com/pipecat-ai/pipecat"
 Website = "https://pipecat.ai"
 
 [project.optional-dependencies]
-anthropic = [ "anthropic~=0.45.2" ]
+anthropic = [ "anthropic~=0.47.2" ]
 assemblyai = [ "assemblyai~=0.36.0" ]
 aws = [ "boto3~=1.35.99" ]
 azure = [ "azure-cognitiveservices-speech~=1.42.0"]
@@ -56,7 +53,7 @@ elevenlabs = [ "websockets~=13.1" ]
 fal = [ "fal-client~=0.5.6" ]
 fish = [ "ormsgpack~=1.7.0", "websockets~=13.1" ]
 gladia = [ "websockets~=13.1" ]
-google = [ "google-cloud-speech~=2.31.0", "google-cloud-texttospeech~=2.25.0", "google-genai~=1.2.0", "google-generativeai~=0.8.4" ]
+google = [ "google-cloud-speech~=2.31.0", "google-cloud-texttospeech~=2.25.0", "google-genai~=1.3.0", "google-generativeai~=0.8.4" ]
 grok = []
 groq = []
 gstreamer = [ "pygobject~=3.50.0" ]

src/pipecat/services/google/google.py

@@ -722,7 +722,9 @@ class GoogleLLMContext(OpenAILLMContext):
 
         self.add_message(glm.Content(role="user", parts=parts))
 
-    def add_audio_frames_message(self, *, audio_frames: list[AudioRawFrame], text: str = None):
+    def add_audio_frames_message(
+        self, *, audio_frames: list[AudioRawFrame], text: str = "Audio follows"
+    ):
         if not audio_frames:
             return
 
@@ -731,8 +733,9 @@ class GoogleLLMContext(OpenAILLMContext):
 
         parts = []
         data = b"".join(frame.audio for frame in audio_frames)
-        if text:
-            parts.append(glm.Part(text=text))
+        # NOTE(aleix): According to the docs only text or inline_data should be needed.
+        # (see https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/inference)
+        parts.append(glm.Part(text=text))
         parts.append(
             glm.Part(
                 inline_data=glm.Blob(