Hume AI's new voice that truly understands emotion
Hume AI's EVI 3 is a new speech-language model for highly expressive, realistic & emotionally intelligent voice AI. Generates any voice/personality from prompts. Outperforms GPT-4o in empathy & naturalness.
Most voice AI today, even the good ones, can still feel a bit generic. They might sound clear, but often lack true personality or a deep sense of emotional connection. Hume AI is tackling this head-on with their new third-gen speech-language model, EVI 3.
EVI 3 offers truly personalizable voice AI. It isn't stuck with a handful of standard voices, it can generate any voice and personality you describe with a prompt. This means interactions can feel much more unique and expressive, as the same underlying intelligence handles understanding you and crafting a spoken response.
Hume AI's blind comparison results suggest it delivers on that human-like quality, with EVI 3 rated higher than GPT-4o on empathy, expressiveness, and naturalness. It also has a neat "fast and slow thinking" ability, connecting to reasoning models and web search as it talks.
You can try EVI 3 now through their live demo and an iOS app (API access is coming). It's a compelling step towards voice AI feeling less like a tool and more like a genuine interaction.
About EVI 3: Understand and generate any voice on Product Hunt
“Hume AI's new voice that truly understands emotion”
EVI 3: Understand and generate any voice launched on Product Hunt on June 3rd, 2025 and earned 162 upvotes and 11 comments, placing #14 on the daily leaderboard. Hume AI's EVI 3 is a new speech-language model for highly expressive, realistic & emotionally intelligent voice AI. Generates any voice/personality from prompts. Outperforms GPT-4o in empathy & naturalness.
On the analytics side, EVI 3: Understand and generate any voice competes within Artificial Intelligence and Audio — topics that collectively have 468.7k followers on Product Hunt. The dashboard above tracks how EVI 3: Understand and generate any voice performed against the three products that launched closest to it on the same day.
Who hunted EVI 3: Understand and generate any voice?
EVI 3: Understand and generate any voice was hunted by Zac Zuo. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Reviews
EVI 3: Understand and generate any voice has received 12 reviews on Product Hunt with an average rating of 5.00/5. Read all reviews on Product Hunt.
For a complete overview of EVI 3: Understand and generate any voice including community comment highlights and product details, visit the product overview.
Zac Zuo
Hi everyone!
Most voice AI today, even the good ones, can still feel a bit generic. They might sound clear, but often lack true personality or a deep sense of emotional connection. Hume AI is tackling this head-on with their new third-gen speech-language model, EVI 3.
EVI 3 offers truly personalizable voice AI. It isn't stuck with a handful of standard voices, it can generate any voice and personality you describe with a prompt. This means interactions can feel much more unique and expressive, as the same underlying intelligence handles understanding you and crafting a spoken response.
Hume AI's blind comparison results suggest it delivers on that human-like quality, with EVI 3 rated higher than GPT-4o on empathy, expressiveness, and naturalness. It also has a neat "fast and slow thinking" ability, connecting to reasoning models and web search as it talks.
You can try EVI 3 now through their live demo and an iOS app (API access is coming). It's a compelling step towards voice AI feeling less like a tool and more like a genuine interaction.