Vocode
Vocode is an open source Python library for building real-time, voice-based LLM applications. It provides abstractions and integrations for transcription, LLM conversation, and speech synthesis, enabling developers to create conversational agents that can interact over phone calls, Zoom, or system audio. It is designed for developers who want to quickly build voice AI apps without managing complex audio pipelines.
✨ Key features
- Real-time streaming conversations with LLMs
- Deploy to phone calls, Zoom, and system audio
- Integrations with multiple transcription, LLM, and synthesis services
- Easy abstractions for building voice agents
- Open source with community contribution
- React SDK available for frontend integration
🎯 Use cases
- Build a voice assistant that talks over your computer's microphone and speaker
- Set up an inbound phone number that answers with an LLM-based agent
- Make outbound phone calls from an LLM-based agent
- Dial into a Zoom meeting with a voice agent
- Integrate voice calls into a Langchain agent
📦 Installation
🧰 Requirements: Python environment with pip; API keys for OpenAI, Azure Speech, and Deepgram (or other services) are required for the quickstart.
pip install vocode
🚀 Usage
import asyncio
import signal
from pydantic_settings import BaseSettings, SettingsConfigDict
from vocode.helpers import create_streaming_microphone_input_and_speaker_output
from vocode.logging import configure_pretty_logging
from vocode.streaming.agent.chat_gpt_agent import ChatGPTAgent
from vocode.streaming.models.agent import ChatGPTAgentConfig
from vocode.streaming.models.message import BaseMessage
from vocode.streaming.models.synthesizer import AzureSynthesizerConfig
from vocode.streaming.models.transcriber import (
DeepgramTranscriberConfig,
PunctuationEndpointingConfig,
)
from vocode.streaming.streaming_conversation import StreamingConversation
from vocode.streaming.synthesizer.azure_synthesizer import AzureSynthesizer
from vocode.streaming.transcriber.deepgram_transcriber import DeepgramTranscriber
configure_pretty_logging()
class Settings(BaseSettings):
openai_api_key: str = "ENTER_YOUR_OPENAI_API_KEY_HERE"
azure_speech_key: str = "ENTER_YOUR_AZURE_KEY_HERE"
deepgram_api_key: str = "ENTER_YOUR_DEEPGRAM_API_KEY_HERE"
azure_speech_region: str = "eastus"
model_config = SettingsConfigDict(env_file=".env", env_file_encoding="utf-8", extra
❓ FAQ
What services does Vocode integrate with for transcription?
Vocode supports AssemblyAI, Deepgram, Gladia, Google Cloud, Microsoft Azure, RevAI, Whisper, and Whisper.cpp.
Can I use Vocode to make phone calls?
Yes, Vocode can set up inbound and outbound phone calls with LLM-based agents, as described in the telephony quickstarts.
Does Vocode provide a React SDK?
Yes, there is a separate React SDK available at https://github.com/vocodedev/vocode-react-sdk.
How can I contribute to Vocode?
Vocode is open source and welcomes contributors. You can refer to the Contribution Guide and Roadmap linked in the README, and join the Discord community.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.