Alice Baird

Research scientist @hume_ai, PhD from @uni__augsburg - audio, speech, paralinguistics, affective computing, wellbeing.

[Alice Baird @Aliceebaird]

https://t.co/xxP6dm66Wf

[Hume AI @hume_ai]

Today, voice models have no problem generating “angry” or “sad” expressions.

But ask for:
→ bored + fast
→ joy + shy
→ disappointment + confident
…and most systems collapse into stereotypes.

Our latest research blog explores why this happens — and how disentangling emotion from voice at the data layer improves expressive control. Read more below!

[Alice Baird @Aliceebaird]

Excited to be co-organizing the ACII 2026 DaiKon Workshop & Challenge! Reach out to register a team: [email protected]

[Hume AI @hume_ai]

We’re excited to launch the 2026 ACII Dyadic Contest (DaiKon) Workshop & Challenge—a new benchmark for modeling emotional influence in dyadic dialogue.

Explore a sample of our conversational audio dataset: 945 sessions, 743 hours, across 5 languages.

Submissions due May 25. We look forward to your participation!

[Alice Baird @Aliceebaird]

Octave 2, now multilingual, so huge! 🥳🔥

[Hume AI @hume_ai]

Introducing Octave 2: our next-generation multilingual text-to-speech model

What’s new:
- Fluent in 11+ languages
- 40% faster (<200ms latency⁠⁠) & 50% cheaper than Octave 1
- Multi-speaker conversation
- More reliable pronunciation
- New voice conversion & phoneme editing capabilities

For the month of October, we’re offering 50% off our Creator plan - use code OCTAVE2 at checkout!

[Hume AI @hume_ai]

Use OpenAI's new open source model with any voice, cloned or designed!

[Hume AI @hume_ai]

2024: Voice Cloning
2025: What about personality cloning?

Hume’s voice AI can now not only mimic your voice but also speaking style and language.

It’s now available via our TTS and new speech-to-speech model, EVI 3, which is also launching today.

[Alice Baird @Aliceebaird]

@hume_ai 🔥 so good!

[Linus ✦ Ekenstam @LinusEkenstam]

This had me floored

This isn’t just a talking model. It understands and expresses voice like a human, across any accent, tone, or style.

It doesn’t just speak. It performs.

Nervous stammer? Confident debate? A whispered secret? It's all in there...

[Alice Baird @Aliceebaird]

Huge! 💚

[Hume AI @hume_ai]

Meet EVI 3, another step toward general voice intelligence.

EVI 3 is a speech-language model that can understand and generate any human voice, not just a handful of speakers. With this broader voice intelligence comes greater expressiveness and a deeper understanding of tune, rhythm, timbre, and speaking style.

[VentureBeat @VentureBeat]

Hume launches new text-to-speech model Octave that generates custom AI voices with adjustable emotions https://t.co/DbFejU2awV

[Alice Baird @Aliceebaird]

Wow, an incredible achievement by the entire team at Hume! 💚

[Hume AI @hume_ai]

Today, we’re releasing Octave: the first LLM built for text-to-speech.

🎨Design any voice with a prompt
🎬 Give acting instructions to control emotion and delivery (sarcasm, whispering, etc.)
🛠️Produce long-form content on our Creator Studio

Unlike traditional TTS that just “reads” words aloud, Octave understands how meaning affects delivery to generate emotional, human-like speech.

[Hume AI @hume_ai]

Octave’s first-of-its-kind voice intelligence speaks for itself 😉 In a blind study, Octave outperformed ElevenLabs Voice Design:

🔊71.6% preferred Octave’s audio quality
🗣️51.7% found Octave more natural
🎯57.7% said Octave better matched voice descriptions

[Hume AI @hume_ai]

Today, we’re releasing Octave: the first LLM built for text-to-speech.

🎨Design any voice with a prompt
🎬 Give acting instructions to control emotion and delivery (sarcasm, whispering, etc.)
🛠️Produce long-form content on our Creator Studio

Unlike traditional TTS that just “reads” words aloud, Octave understands how meaning affects delivery to generate emotional, human-like speech.

[Hume AI @hume_ai]

Better AI voices coming soon...

[Hume AI @hume_ai]

Introducing OCTAVE, a next-generation speech-language model.

OCTAVE has new emergent capabilities, like on-the-fly voice and personality creation and much more 👇

[Hume AI @hume_ai]

Hume's EVI 2 generates speech and language in tandem, with rich context and emotion. It is the only speech-LLM model on the market that seamlessly integrates with frontier LLMs like @AnthropicAI's Claude 3.5 Sonnet.

Together, EVI 2 + Claude 3.5 Sonnet:
🎙️Over 2 million minutes of voice AI conversations completed

✨36% of users choose Claude, higher than any other LLM integrated with EVI 2

💸Developers using EVI 2 + Claude see an 80% reduction in cost and 10% decrease in latency through prompt caching

[Hume AI @hume_ai]

Loneliness and cognitive decline are major health challenges for older adults. https://t.co/KroZsgBrbo integrated Hume’s Empathic Voice Interface (EVI) into their digital companionship app, with powerful results:

🧠88% reported increased mental stimulation
🫂90% experienced reduced loneliness

All from just a few 15-minute sessions over 5 weeks.

Read the case study: https://t.co/sEcebD78Rt

[Hume AI @hume_ai]

EVI 2 is available now 👉 API https://t.co/Kw2o7HVgoR &amp; live demo https://t.co/0zBH3qR39D

[Hume AI @hume_ai]

Thanks for taking the time to chat with EVI 2 @willknight → https://t.co/0zBH3qRAZb

[Alice Baird @Aliceebaird]

So proud of our team! See what we’ve been up to: https://t.co/STDnqXp7hD

[Hume AI @hume_ai]

Introducing Empathic Voice Interface 2 (EVI 2), our new voice-to-voice foundation model. EVI 2 merges language and voice into a single model trained specifically for emotional intelligence.

You can try it and start building today.

[Hume AI @hume_ai]

New Hume publication!

What’s the relationship between facial expressions and reported feelings, and how does it vary across cultures?

We gathered the most definitive evidence to date:

45,231 recordings of participants from North America, Europe, and Japan, reacting to 2,185 videos and reporting on their experiences.

Here’s what we found 👇