Vocode

Open-source platform for building and scaling hyperrealistic voice AI agents with enterprise solutions.

Last verified:

Visit Vocode

What is Vocode?

Vocode is an open-source voice AI library that makes it easy to build voice-based LLM applications. Using Vocode, you can create real-time streaming conversations with LLMs and deploy them to phone calls, Zoom meetings, personal assistants, and voice-based apps. The platform provides both a hosted API service for enterprise-grade phone call management and an open-source core library for self-hosted deployments.

Vocode Core provides the integrations, orchestration, and abstractions to create voice applications on top of any AI stack. It includes conversation abstractions for streaming and turn-based interactions, conversation functionality like endpointing and emotion tracking, integrations with top speech-to-text and text-to-speech providers, and cross-platform support for telephony, web, and Zoom. Vocode API is an enterprise-grade API to manage AI agents on phone calls that lives on top of Vocode Core.

The tool is designed for developers building voice AI applications, teams wanting to automate inbound/outbound phone calls, companies creating personal assistants, and organizations deploying AI call centers. Users can program agents to execute arbitrary tasks, embed AI call centers directly in product operations flows, set up LLMs to answer/make phone calls, act as personal assistants, or join Zoom meetings.

Key features include Python and Node.js SDKs, guardrails for reliable agent performance, metrics for analyzing performance, conversation analytics, call recording, spam protection, monitoring and logging, multilingual bot support, external actions, warm transfer capabilities, IVR navigation, vector database integration, and HIPAA compliance options.

Vocode pricing

Pricing model: Freemium

Vocode offers three main plans: Free ($0/month) includes basic voices and models, unlimited conversation minutes (limited time), API access, conversation analytics, and inbound calls only. Developer Plan ($25/month or $288/year) includes advanced voices and models, 100 minutes of premium conversation time/month, 3 included phone numbers, API access with external actions, and inbound only. Enterprise Plan (contact support for pricing) includes dedicated support and infrastructure, white glove service, custom minutes, custom phone numbers, custom models and voices, conversation analytics, inbound and outbound calls (via request), self-hosted and hosted options, 24/7 dedicated support, and enterprise SLAs. Individual phone numbers cost $3/month or $30/year with spam protection, monitoring and logging, conversation analytics, and call recording.

Vocode pros

  • Open source library free to use
  • Python and Node.js SDKs available
  • Real-time streaming conversations with LLMs
  • Deploy to phone calls, Zoom, web
  • Integrations with top STT/TTS providers
  • Enterprise-grade API for phone calls
  • Conversation analytics included
  • Call recording capability
  • Multilingual bot support
  • Guardrails for reliable performance
  • Metrics to analyze performance
  • Customizable prompts and voices
  • External actions support
  • Warm transfer beta feature
  • IVR navigation beta feature
  • Vector databases integration
  • HIPAA compliance option
  • Cross-platform telephony support
  • Endpointing and emotion tracking
  • Configurable interruption sensitivity

Vocode cons

  • Free plan limited to inbound calls only
  • Developer plan only 100 premium minutes/month
  • Outbound calls require enterprise request
  • Individual phone numbers cost $3/month
  • Enterprise pricing not publicly listed
  • Some features still in beta
  • Requires API key for hosted services
  • Self-hosting requires technical setup
  • Limited concurrent calls on lower tiers
  • Premium voices on paid plans only

Frequently asked questions about Vocode

What is Vocode?

Vocode is an open-source library for building voice-based LLM applications. It provides tools and abstractions to build any kind of voice-based application on top of LLMs, including setting up LLMs to answer/make phone calls, act as personal assistants, join Zoom meetings, and more. Vocode offers both a hosted API service and an open-source core library.

What are the main components of Vocode?

Vocode has two main components: Vocode Core provides the integrations, orchestration, and abstractions to create voice applications on top of any AI stack. Vocode API is an enterprise-grade API to manage AI agents on phone calls that lives on top of Vocode Core.

What SDKs does Vocode provide?

Vocode provides Python SDK (vocode-api package) and Node.js SDK (@vocode/vocode-api package). You can install Python SDK using poetry add vocode-api and Node SDK using npm install --save @vocode/vocode-api.

What can I build with Vocode?

You can build real-time streaming conversations with LLMs, deploy to phone calls and Zoom meetings, create personal assistants, build voice-based apps like voice-based chess, automate inbound and outbound phone calls, embed AI call centers in product operations flows, and set up LLMs to answer/make phone calls or act as personal assistants.

What integrations does Vocode support?

Vocode provides integrations to all of the best speech-to-text and text-to-speech providers. It supports cross-platform integration for telephony, web, and Zoom platforms.

What conversation features does Vocode provide?

Vocode provides conversation abstractions for streaming and turn-based interactions, conversation functionality including endpointing and emotion tracking, configurable interruption sensitivity through Conversational Dials, and support for both streaming responses and turn-based conversations.

What are agents in Vocode?

Vocode Agents are the core component of the Vocode API. Agents can be deployed to receive phone calls by attaching them to Numbers, or used to make outbound calls via the calls/create endpoint. An agent is an amalgamation of Prompts (instructions controlling behavior), Voices (synthetic voice), Webhooks (event subscription mechanism), Conversational Dials (communication controls), and Actions (tools like ending conversation or transferring calls).

What beta features does Vocode offer?

Vocode offers several beta features including External Actions, Warm Transfer, Multilingual bots, Injecting context, IVR Navigation, Vector databases, and HIPAA Compliance.

How do I get started with Vocode hosted service?

To get started with Vocode hosted service, install either the Python or Node SDK, or hit the API directly. Initialize the Vocode client with your API key token. You can get an API key from app.vocode.dev. Then go through the Hosted Walkthrough which shows how to start interacting with the API, including setting up your number, configuring it, and setting up webhooks.

What is the difference between hosted and open source Vocode?

The hosted service allows you to automate inbound and outbound phone calls via Vocode's API and SDKs with scalable, production-ready infrastructure, guardrails, and metrics. The open source library (Vocode Core) is free to use and self-hosted, providing integrations, orchestration, and abstractions to create voice applications on top of any AI stack with maximum developer flexibility.

Categories

Use cases

Browse all AI tools on NeedAnAI