# Hume EVI Review

> Hume EVI review: Emotionally intelligent speech-to-speech API (Empathic Voice Interface, EVI 3 / EVI 4 mini) with voice cloning and supplemental-LLM support.

Emotionally intelligent speech-to-speech API (Empathic Voice Interface, EVI 3 / EVI 4 mini) with voice cloning and supplemental-LLM support.

Among the 17 voice ai tools we track, Hume EVI has the 6th-lowest latency - a fit for real-time, conversational apps.

From $0.07/min · facts verified Jul 30, 2026

## What we know about Hume EVI

Hume EVI sits in the voice ai category, where it is emotionally intelligent speech-to-speech API (Empathic Voice Interface, EVI 3 / EVI 4 mini) with voice cloning and supplemental-LLM support. We keep this Hume EVI profile grounded in primary sources, each fact dated to when we last confirmed it.

On pricing, Hume EVI starts at $0.07 per minute for its entry tier. That is the sticker rate: real production cost usually runs higher once you add a language model, a voice provider, and telephony minutes.

New Hume EVI accounts get Free: 5 EVI min + 10k TTS chars/mo to test with before a card is required, which is enough to run a handful of real calls and judge latency and voice quality for your own script rather than a demo.

On capabilities, Hume EVI covers bring-your-own llm, bring-your-own tts voice, interruption handling (barge-in), consent/recording compliance tooling, no-code agent builder, and api-first (full lifecycle via api), and does not offer annual contract required for best pricing?. Each of those is verified against Hume EVI's own docs or dashboard, not marketing copy.

On performance, Hume EVI reports a median end-to-end latency around 300 ms. Latency figures are vendor claims until we measure them ourselves, and they are labeled that way on the Hume EVI fact sheet, so treat them as a starting point rather than a guarantee.

For compliance, with Hume EVI: HIPAA is included on the plans we checked. If you are in a regulated space, confirm the current posture with Hume EVI before you commit, since these change plan by plan.

Hume EVI ranks the 6th-lowest latency of the 17 voice ai tools we track, but only 23rd of 17 for languages, so where it lands for you depends on which of those matters more.

In total we track 15 verified facts for Hume EVI today, each linking the primary source it came from so you can check our work - and vendor claims we have not measured ourselves are labeled as such on the Hume EVI fact sheet below.

## Hume EVI pricing

≈ $0.07 per minute all-in (est.) plus $3/mo platform fee, verified Jul 30, 2026 ([source](https://www.hume.ai/pricing)).

| Monthly volume | All-in monthly bill |
| --- | --- |
| 1K min/mo | $73 |
| 10K min/mo | $703 |
| 100K min/mo | $7,003 |

All-in monthly: platform fee + per-minute + telephony. Assumes GPT-4o class model, ElevenLabs voice.

## Fact sheet

### Pricing

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Base price / min | $0.07 | Jul 30 | [source](https://www.hume.ai/pricing) |
| Free tier credit | Free: 5 EVI min + 10k TTS chars/mo | Jul 30 | [source](https://www.hume.ai/pricing) |
| Monthly platform fee | $3 | Jul 30 | [source](https://www.hume.ai/pricing) |
| Annual contract required for best pricing? | ✗  No | Jul 30 | [source](https://www.hume.ai/pricing) |

### Capabilities

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Median e2e latency | ~300 ms | Jul 30 | [source](https://www.hume.ai/empathic-voice-interface) |
| Bring-your-own LLM | ✓  Yes | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/overview) |
| Bring-your-own TTS voice | ✓  Yes | Jul 30 | [source](https://www.hume.ai/empathic-voice-interface) |
| Languages supported | 11 count | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/overview) |
| Interruption handling (barge-in) | ✓  Yes | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/overview) |

### Compliance & trust

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| HIPAA BAA on base plan | ✓  Included | Jul 30 | [source](https://dev.hume.ai/docs/resources/privacy) |
| Consent/recording compliance tooling | ✓  Yes | Jul 30 | [source](https://dev.hume.ai/docs/resources/privacy) |

### Build experience

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| No-code agent builder | ✓  Yes | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/features/tool-use) |
| API-first (full lifecycle via API) | ✓  Yes | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/overview) |
| Native integrations count (+ key ones: GHL, HubSpot, Zapier, Make, Twilio) | Claude, GPT, Gemini + BYO LLM; 5 SDKs | Jul 30 | [source](https://dev.hume.ai/docs/speech-to-speech-evi/features/tool-use) |

### Commercial

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Self-serve signup | ✓  Yes | Jul 30 | [source](https://www.hume.ai/pricing) |

## Hume EVI head-to-head

| Comparison | Record | Per use case |
| --- | --- | --- |
| [Hume EVI vs Vapi](https://www.versusref.com/voice-ai/hume-evi-vs-vapi/) | won 1 · lost 5 · tied 0 | No-code agency deploys: lost; Custom pipeline control: lost; Ecommerce support and orders: lost; Healthcare compliance: won; Outbound sales campaigns: lost; SMB receptionist and booking: lost |
| [Hume EVI vs OpenAI Realtime API](https://www.versusref.com/voice-ai/hume-evi-vs-openai-realtime-api/) | won 4 · lost 0 · tied 2 | No-code agency deploys: won; Custom pipeline control: won; Ecommerce support and orders: tie; Healthcare compliance: won; Outbound sales campaigns: tie; SMB receptionist and booking: won |

Source: https://www.versusref.com/voice-ai/tools/hume-evi/