Headquarters
New York, NYFounded
2021CEO / Founder
Alan CowenEmployees
51-200Funding
Series B — $77MProducts
1 product listedAvg. AI Score
7.9/10Website
hume.ai/Profile updated
Hume AI is an AI research company founded in 2021 by Alan Cowen and headquartered in New York, NY. The company focuses on building models that interpret human emotional and expressive behavior from voice, facial expressions, and text, with the stated goal of aligning AI outputs with human well-being.
Hume's primary product is the Empathic Voice Interface (EVI), a conversational AI API that incorporates vocal tone and emotional cues into real-time dialogue. The company also offers an Expression Measurement API that analyzes facial, vocal, and linguistic signals across dozens of emotional dimensions. These tools are used by developers and enterprises building applications in mental health, customer experience, and human-computer interaction.
Hume AI raised a $50 million Series B round in 2024, bringing its total funding to approximately $77 million. Earlier backers include Union Square Ventures, Nat Friedman, and Daniel Gross. The company employs between 51 and 200 people and has published peer-reviewed research on emotional AI, drawing on Cowen's prior academic work at UC Berkeley on the taxonomy of human emotion.
Dated, source-cited figures. Estimates are marked.
All products by Hume AI reviewed by our AI panel of experts.
Six independent reviewer personalities assessed Hume AI's product on their own terms. They disagreed by 1.0 points — The Skeptic was the hardest to convince, The Domain Strategist the most positive.
Averaged across 1 reviewed product in our catalog. Scores are out of 10.
Who Hume AI competes with, and how they differ.
ElevenLabs leads on breadth — text-to-speech, voice cloning, dubbing, and an agent platform serving media/localization buyers — at a far larger scale than Hume's narrower focus on emotion measurement and empathic response generation grounded in psychology research.
1 product reviewed on TopReviewedCartesia's Sonic models target the same low-latency speech-to-speech niche as EVI, competing on state-space-model efficiency and sub-100ms responses, but without Hume's emotion-research base or prosody-driven response tuning.
1 product reviewed on TopReviewedOpenAI's Realtime API and ChatGPT Advanced Voice Mode compete for the same conversational voice-agent developers with far greater distribution and lower list prices, without Hume's dedicated emotional-expression research or granular vocal-affect controls.
2 products reviewed on TopReviewedDeepgram competes on enterprise speech-to-text/text-to-speech infrastructure and volume pricing for call centers and transcription pipelines, rather than an emotion-aware conversational layer.
1 product reviewed on TopReviewedAssemblyAI focuses on speech-to-text and audio intelligence (summarization, sentiment tagging) for analytics pipelines — a different buyer than Hume's real-time empathic-conversation API.
1 product reviewed on TopReviewedBland AI targets automated phone-call agents for sales and support workflows, competing for voice-agent budget but without Hume's emotion-modeling research or its Octave/EVI model stack.
1 product reviewed on TopReviewedKey milestones in Hume AI's history, with sources.
Google DeepMind licensing deal; founder Alan Cowen departs as CEO
Google DeepMind entered a non-exclusive technology licensing agreement with Hume AI and hired CEO/founder Alan Cowen along with roughly seven engineers to work on Gemini's voice features. Andrew Ettinger, an investor and executive who had recently joined Hume, became Hume AI's new CEO; the company said it would continue serving existing enterprise clients under the licensing arrangement.
techcrunch.comOctave 2 and EVI 4 mini launched
Hume launched Octave 2, a faster and cheaper text-to-speech model (under 200ms, half the price of Octave 1), and EVI 4 mini, bringing Octave 2's capabilities to Hume's speech-to-speech API across 11 languages.
hume.aiEVI 3 introduced
Hume introduced EVI 3, its third-generation speech-language foundation model, with improved expressiveness, realism, and emotional understanding; made available via live demo and iOS app, with API access announced July 17, 2025.
hume.aiEVI 2 launched in beta
Hume released EVI 2, its second-generation voice-to-voice foundation model, in public beta via its app and API.
hume.aiEmpathic Voice Interface (EVI) unveiled
Hume unveiled its Empathic Voice Interface (EVI), described as a first-of-its-kind conversational AI with emotional intelligence, announced alongside its Series B round; general API access followed on April 18, 2024.
hume.aiSeries B — $50M led by EQT Ventures
Hume AI raised a $50 million Series B led by EQT Ventures, with participation from Union Square Ventures, Nat Friedman & Daniel Gross, Metaplanet, Northwell Holdings, Comcast Ventures, and LG Technology Ventures.
hume.aiSeries A — $12.7M led by Union Square Ventures
Hume AI raised a $12.7 million Series A led by Union Square Ventures, with participation from Northwell Holdings, Comcast Ventures, LG Technology Ventures, Wisdom Ventures, and Pinterest co-founder Evan Sharp, to expand its emotion-detection technology beta.
axios.comHume AI founded by Alan Cowen
Hume AI was founded in March 2021 by Alan Cowen, a computational emotion scientist with a PhD in psychology from UC Berkeley who previously led the Affective Computing team at Google AI, to build voice AI models that measure and respond to human emotional expression.
research.contrary.comBrowse multi-perspective AI panel reviews across hundreds of AI tools, agents, and platforms. Find the right software with insights from CTO, Developer, Marketer, Finance, and User perspectives.