Vapi
Platform for building voice agents
Development platform for creating, testing and deploying voice agents using artificial intelligence. Enables making and receiving telephone calls, integrating voice conversations into web applications and connecting with external systems through APIs. Supports over 100 languages.
585
Social networks of Vapi
¡Producto agotado!
Favorite
Sign up and save your AI tools!
RegistrationWhat is Vapi?
Platform for developers that provides API-first infrastructure for building voice agents using artificial intelligence. The system enables creating voice assistants that can handle telephone calls, customer support, bookings and other operations through conversation.
Each voice agent combines three technologies: speech-to-text conversion, language model for processing and generating responses, and text-to-speech conversion. The platform allows selecting from multiple providers and models, including OpenAI, Anthropic, Google, Deepgram and ElevenLabs, among others.
Vapi offers two structures for building agents: Assistants, based on a single system prompt with tools and structured outputs, suitable for cases such as customer support or lead qualification; and Squads, which orchestrate multiple specialised assistants with context-preserving transfers, useful for configurations requiring specialisation such as medical triage or e-commerce management.
The platform includes telephone integration capabilities for inbound and outbound calls, web integration through SDK, and connection with APIs and external systems via tool calling. Developers have SDKs in TypeScript, Python and React, along with CLI for terminal development. Response times are below 500 milliseconds with 99.99% uptime. The system supports over 100 languages and allows bringing your own models through API keys or self-hosted models.
Vapi provides automated testing tools to identify hallucination risks before production, A/B experiments to optimise prompts and flows, and performance analytics. The platform includes conversational guardrails to prevent model hallucinations. For enterprise implementations, it offers SOC2, HIPAA and PCI certifications.
Each voice agent combines three technologies: speech-to-text conversion, language model for processing and generating responses, and text-to-speech conversion. The platform allows selecting from multiple providers and models, including OpenAI, Anthropic, Google, Deepgram and ElevenLabs, among others.
Vapi offers two structures for building agents: Assistants, based on a single system prompt with tools and structured outputs, suitable for cases such as customer support or lead qualification; and Squads, which orchestrate multiple specialised assistants with context-preserving transfers, useful for configurations requiring specialisation such as medical triage or e-commerce management.
The platform includes telephone integration capabilities for inbound and outbound calls, web integration through SDK, and connection with APIs and external systems via tool calling. Developers have SDKs in TypeScript, Python and React, along with CLI for terminal development. Response times are below 500 milliseconds with 99.99% uptime. The system supports over 100 languages and allows bringing your own models through API keys or self-hosted models.
Vapi provides automated testing tools to identify hallucination risks before production, A/B experiments to optimise prompts and flows, and performance analytics. The platform includes conversational guardrails to prevent model hallucinations. For enterprise implementations, it offers SOC2, HIPAA and PCI certifications.