Hume AI
Hume AI is a research lab and technology company building the data, models, benchmarks, evaluation systems, and feedback infrastructure for emotionally intelligent voice AI. Its stated mission is to ensure AI is built to serve human goals and emotional well-being.
At a Glance
- Frontier AI labs and foundation-model developers
- AI-first enterprises and enterprise software teams
- Contact centers and customer-service providers
- Healthcare providers, digital health, and mental-health companies
- +4 more
AI Tools by Hume AI
(1)TADA
Open Source Text to Speech Model
Discussions
No discussions yet
Be the first to start a discussion about Hume AI
Latest News
Hume introduced the Audio Decisions API, an audio-native decision layer for structured judgments about speech, sound, and context.
Hume launched Expression APIs for real-time measurement of facial and vocal expression.
Hume introduced the Voice Controllability Leaderboard to evaluate whether TTS models follow expressive directions.
Hume open-sourced TADA, a fast speech-generation framework based on Text-Acoustic Dual Alignment, with code and pretrained models.
Products & Services
A conversational voice-to-voice platform combining transcription, an LLM, expression measures, and expressive speech. It supports end-of-turn detection, interruption handling, response to vocal expression, expressive TTS, and application-specific optimization.
A foundational voice-to-voice model focused on natural turn-taking, expressive speech, and emotionally responsive conversation.
Customizable speech-to-speech model/API with hyperrealistic voice cloning and integrations including Claude 4, Gemini 2.5, and Kimi K2.
Omni-Capable Text and Voice Engine for expressive text-to-speech, prompt-created voices, personality creation, and emotion-directed speech.
Market Position
Hume positions itself as the emotional-intelligence, data, and evaluation layer for voice AI rather than only a voice generator or generic LLM. Its differentiation is scientifically grounded expression data, human preference measurement, expressive speech models, and evaluation/post-training infrastructure spanning the full voice-model lifecycle. Competitors include ElevenLabs, OpenAI voice models, Google Gemini/Cloud speech products, PlayHT, Cartesia, Deepgram, and other voice-agent and TTS platforms.
Leadership
Founders
Alan Cowen
Psychologist and AI researcher with a PhD from UC Berkeley; previously led the Affective Computing team at Google AI. His research on Semantic Space Theory underpins Hume's approach to measuring human expression.
John Beadle
Co-founder and managing partner of Aegis Ventures; founding investor and board member of Hume AI, with a background in venture building and healthcare technology.
Executive Team
Andrew Ettinger
Chief Executive Officer and Board Member
AI and data-infrastructure executive with 15 years of leadership experience; previously Chief Revenue Officer at Appen and held leadership roles at Pivotal and Astronomer.
John Beadle
Chief Financial Officer and Board Member
Aegis Ventures co-founder and managing partner, venture investor, and Hume co-founder.
Board of Directors
Founding Story
Alan Cowen founded Hume after studying human emotion and becoming concerned that voice assistants and language models lacked sensitivity to emotion, human preferences, and well-being. Hume was started to develop scientifically grounded AI that could understand vocal, facial, and linguistic expression and optimize interactions for human well-being rather than superficial engagement metrics.
Business Model
Revenue Model
Developer and enterprise APIs, usage-based voice generation and expression/decision inference, enterprise research partnerships, custom data and annotation services, model evaluation, post-training, and reinforcement-learning infrastructure.
Target Markets
- Frontier AI labs and foundation-model developers
- AI-first enterprises and enterprise software teams
- Contact centers and customer-service providers
- Healthcare providers, digital health, and mental-health companies
- Education and research institutions
- Automotive, robotics, hardware, and consumer-device companies
- Voice agents and customer support
- Healthcare, mental-health coaching, and patient screening/triage
- Education, language learning, and interview preparation
- Robotics, automotive assistants, and consumer devices
- Audiobooks, podcasts, games, avatars, and creative voice production
- Financial communications and earnings-call analysis
- Markets EQ
- Northwell Health Holdings
- hpy
- Everfriends