MODELS

EVI 4-mini

by Hume AI

modelvoice-aievi

Overview

Multilingual real-time speech-language model for emotionally intelligent voice interfaces with streaming speech and prosody-aware responses.

Details

EVI 4-mini is a Hume AI speech-to-speech model for the Empathic Voice Interface. Hume’s official documentation describes EVI as real-time emotionally intelligent voice AI and lists EVI 4-mini in its version comparison. Hume’s developer docs state that EVI 4-mini is live with expanded language support and lower latency for faster, more natural responses. The model is positioned for multilingual, real-time voice interactions with streaming speech and prosody-aware responses.

When to Use

Build real-time conversational voice interfaces that need streaming speech-to-speech interaction. Evaluate when multilingual voice support lower latency and more natural responses are important requirements. Use for applications where prosody-aware responses and emotionally intelligent voice behavior are relevant.

Getting Started

Review the EVI overview in Hume’s official developer documentation.
Check Hume’s Speech-to-Speech EVI documentation and version comparison for EVI 4-mini availability and behavior.
Review the pricing page for EVI 4 mini usage rates
included monthly usage
and concurrent connection limits.
Review Hume’s use-case guidelines before using emotion-related model outputs in a commercial application.

Key Features

•Real-time speech-to-speech voice AI through Hume’s Empathic Voice Interface.
•EVI 4-mini is listed in Hume’s official EVI version comparison.
•Expanded language support and lower latency are stated in Hume’s developer documentation.
•Supports streaming speech and prosody-aware responses according to the supplied official source excerpts.

Capabilities

•real-time speech-to-speech
•multilingual voice interaction
•streaming speech
•prosody-aware responses
•emotionally intelligent voice interface

Last updated Jul 31, 2026