217f03b9cc
Thinking, sometimes called "extended thinking" or "reasoning", is an LLM process where the model takes some additional time before giving an answer. It's useful for complex tasks that may require some level of planning and structured, step-by-step reasoning. The model can output its thoughts (or thought summaries, depending on the model) in addition to the answer. The thoughts are usually pretty granular and not really suitable for being spoken out loud in a conversation, but can be useful for logging or prompt debugging. Here's what's added: 1. New typed input parameters for Google and Anthropic LLMs that control the models' thinking behavior (like how much thinking to do, and whether to output thoughts or thought summaries). 2. New frames for representing thoughts output by LLMs. 3. A generic mechanism for associating extra LLM-specific data with a function call in context, used specifically to support Google's function-call-related "thought signatures", which are necessary to ensure thinking continuity between function calls in a chain (where the model thinks, makes a function call, thinks some more, etc.) 4. A generic mechanism for recording LLM thoughts to context, used specifically to support Anthropic, whose thought signatures are expected to appear alongside the text of the thoughts within assistant context messages. 5. An expansion of `TranscriptProcessor` to process LLM thoughts in addition to user and assistant utterances.
Pipecat Foundational Examples
This directory contains examples showing how to build voice and multimodal agents with Pipecat. Each example demonstrates specific features, progressing from basic to advanced concepts.
Setup
-
Follow the README steps to get your local environment configured.
Run from root directory: Make sure you are running the steps from the root directory.
Using local audio?: The
LocalAudioTransportrequires a system dependency forportaudio. Install the dependency to use the transport. -
Copy the
env.examplefile and add API keys for services you plan to use:cp env.example .env # Edit .env with your API keys -
Navigate to the examples directory if you aren't already there:
cd examples/foundational -
Run any example:
uv run python 01-say-one-thing.py -
Open the web interface at http://localhost:7860/client/ and click "Connect"
Running examples with other transports
Most examples support running with other transports, like Twilio or Daily.
Daily
You need to create a Daily account at https://dashboard.daily.co/u/signup. Once signed up, you can create your own room from the dashboard and set the environment variables DAILY_SAMPLE_ROOM_URL and DAILY_API_KEY. Alternatively, you can let the example create a room for you (still needs DAILY_API_KEY environment variable). Then, start any example with -t daily:
uv run 07-interruptible.py -t daily
Twilio
It is also possible to run the example through a Twilio phone number. You will need to setup a few things:
- Install and run ngrok.
ngrok http 7860
- Configure your Twilio phone number. One way is to setup a TwiML app and set the request URL to the ngrok URL from step (1). Then, set your phone number to use the new TwiML app.
Then, run the example with:
uv run 07-interruptible.py -t twilio -x NGROK_HOST_NAME
Examples by Feature
Basics
- 01-say-one-thing.py: Most basic bot that says one phrase and exits (Transport, TTS, Event handlers)
- 02-llm-say-one-thing.py: Bot generates a response with an LLM (LLM initialization)
- 03-still-frame.py: Displays a static image (Video transport, Image service)
- 04-transport.py: Different transport options (WebRTC, Daily, Livekit)
Conversational AI
- 07-interruptible.py: Basic voice assistant bot (STT, TTS, LLM, Interruptible speech)
- 10-wake-phrase.py: Bot activated by wake phrase (WakeCheckFilter)
- 22-natural-conversation.py: Smart turn detection (Multiple LLMs, Turn management)
- 38-smart-turn-fal.py: ML-based turn detection (Fal service, Local models)
Common Utilities
- 17-detect-user-idle.py: Handle inactive users (UserIdleProcessor)
- 24-stt-mute-filter.py: Selectively mute user input (STTMuteFilter)
- 28-transcription-processor.py: Record conversation text (TranscriptProcessor)
- 30-observer.py: Access frame data (Custom observers)
- 31-heartbeats.py: Detect idle pipelines (Pipeline monitoring)
- 34-audio-recording.py: Record conversation audio (Composite and track-level recording)
Advanced LLM Features
- 14-function-calling.py: Bot with tool usage (Function schemas, Tool registration)
- 20a-persistent-context-openai.py: Persistent conversation context (Memory management)
- 32-gemini-grounding-metadata.py: Web search capabilities (Google search integration)
- 33-gemini-rag.py: Retrieval-augmented generation (Data sources, Grounding)
- 37-mem0.py: Long-term agent memory (Mem0 service integration)
Media Handling
- 05-sync-speech-and-images.py: Synchronized narration with images (Custom processors, SyncParallelPipeline)
- 06a-image-sync.py: Dynamic image updates while speaking (Synchronized A/V pipelines)
- 09-mirror.py: Mirror user's audio and video (Custom frame processors)
- 11-sound-effects.py: Add sounds when bot speaks (Sound playback, Event synchronization)
- 23-bot-background-sound.py: Play background audio (SoundfileMixer)
Vision & Multimodal
- 12a-describe-video-gemini-flash.py: Bot describes user's video (Video input, Multimodal LLMs)
- 26c-gemini-live-video.py: Gemini with video input (Streaming video, Function calls)
Voice & Language
- 13-transcription.py: Speech transcription demo (STT providers, Real-time transcription)
- 15-switch-voices.py: Dynamic voice/language changing (ParallelPipelines, FunctionFilters)
- 25-google-audio-in.py: Gemini for speech recognition (Alternative transcription)
- 35-pattern-pair-voice-switching.py: Dynamic TTS voice switching (XML parsing, PatternPairAggregator)
- 36-user-email-gathering.py: Spelling mode for TTS (Confirmation patterns, XML tags)
Integration Examples
- 18-gstreamer-filesrc.py: GStreamer video streaming (Video processing)
- 19-openai-realtime-beta.py: OpenAI Speech-to-Speech (Direct S2S, Function calls)
- 21-tavus-layer-tavus-transport.py: Tavus digital twin (Avatar integration)
- 27-simli-layer.py: Simli avatar integration (Video synchronization)
Performance & Optimization
- 16-gpu-container-local-bot.py: GPU-accelerated local bot (Performance measurement)
Advanced Usage
Customizing Network Settings
uv run python <example-name> --host 0.0.0.0 --port 8080
Troubleshooting
- No audio/video: Check browser permissions for microphone and camera
- Connection errors: Verify API keys in
.envfile - Port conflicts: Use
--portto change the port
For more examples, visit our the `pipecat-examples repository.