HTTP 200 verified twice
Voice AI · Tool
Hume AI
ResearchedHume AI is the human evaluation layer for voice, speech, and conversational AI, offering multimodal emotional intelligence across 50+ languages. It provides evaluation APIs, the Kairos platform, Octave TTS and EVI speech-to-speech models, and SDKs for building empathic voice experiences.
Official site snapshots
Visit live site See the official site at a glance
Read-only public captures of Hume AI’s homepage. Screenshots are dated, never live embeds, and open full-screen.
In one minute
Start here for the decision-making essentials: what Hume AI does, who it is for, how it is accessed, and the first-party sources behind this profile.
Pricing4 options
Free plan at $0/month with 10,000 TTS characters and 5 minutes of EVI usage
Starter plan at $3/month (first month 50% off) with 30,000 TTS characters
Creator plan at $7/month (regularly $14/month) with 140,000 TTS characters
Pro plan at $70/month with 1,000,000 TTS characters
Platforms
TypeScriptReactPythonSwift (iOS and macOS)
API accessNot public
FoundedNot disclosed by source
AvailabilityWeb / remote
LicenseOpen source SDKs, examples, and tools on GitHub
Best suited to
Source-backed fit Voice-native AI teams who need to know what callers are actually feeling Engineers and developers building empathic voice applications Voice AI teams needing evaluation suites and benchmarking infrastructure Organizations requiring SOC 2 Type II, GDPR, and HIPAA compliant AI… Engineers and developers business API customers
Decision support
6 sourced answersCommon questions and adoption checks
Short answers to the questions buyers and builders commonly ask about Hume AI. Each answer cites the shared ledger below, where every source is listed once.
01What does Hume AI say it can do?
Human evaluation layer for voice, speech, and conversational AI · Multimodal emotional intelligence spanning 50+ languages, 48+ emotions, and 600+ voice des · Build and measure AI based on real human judgment, grounded in decades of research · Conversational AI
The human evaluation layer for voice, speech, and conversational AI.
02Who is Hume AI intended for?
Voice-native AI teams who need to know what callers are actually feeling · Engineers and developers · business API customers (governed by separate agreements) · General users of Hume AI website and services (collectively "Offerings")
built for voice-native AI teams who need to know what callers are actually feeling
03What use cases does Hume AI describe?
Custom data collection built around specific scenarios, use cases, and evaluation requirem · Adding voice, expression, and empathy to applications · Bringing stories to life with human-like expression · empathic AI services
Custom data collection built around your specific scenarios, use cases, and evaluation requirements.
04What should teams verify before adopting Hume AI?
Privacy policy does not apply to business API customers, research surveys/studies, or empl · Research surveys and studies are covered by a separate Research Privacy Policy · Privacy policy does not apply to employees or job applicants · References the '30% problem' — fifty years of scientific focus on only six emotions covers
Does not apply to content processed for business API customers (governed by separate agreements)
05What pricing information is available for Hume AI?
Free plan at $0/month with 10,000 TTS characters and 5 minutes of EVI usage · Starter plan at $3/month (first month 50% off) with 30,000 TTS characters · Creator plan at $7/month (regularly $14/month) with 140,000 TTS characters · Pro plan at $70/month with 1,000,000 TTS characters
Free $0 / month Sign up
06Does Hume AI document API access?
Expression Measurement API providing real-time emotion tagging and offline analysis across · Human Feedback API providing pre-screened human raters returning per-sample scores, free-r · Hume API with documentation, tutorials, and API references · Expression Measurement API and Human Feedback API
Real-time emotion tagging and offline analysis across 48+ emotions and 50+ languages with 600+ output metrics
Decision guide
Voice AICapabilities and operating fit
This profile connects the jobs Hume AI is described as handling with its delivery model, access options and the subjects used to match it to related products in this directory.
Common use cases
- Custom data collection built around specific scenarios, use cases, and evaluation requirem
- Building evaluation suites, simulating agent-to-agent and human-to-agent conversations, an
- Real-time emotion tagging and offline analysis across 48+ emotions and 50+ languages with
- Obtaining pre-screened human rater scores, free-response feedback, and aggregated analysis
- Generating expressive speech with natural emotion and customizable voices via Octave TTS
- Transforming voices while preserving emotion via Voice Conversion
Topics mapped
Verified capabilities
- Human evaluation layer for voice, speech, and conversational AI
- Multimodal emotional intelligence spanning 50+ languages, 48+ emotions, and 600+ voice des
- Build and measure AI based on real human judgment, grounded in decades of research
- Conversational AI
- Expressive Text-to-Speech
- Voice Conversion
- Text-to-speech built on a language model that interprets the meaning of text rather than j
- Customizable voices for any character
Intended audiences
Access signals
- Pricing model
- Free plan at $0/month with 10,000 TTS characters and 5 minutes of EVI usage · Starter plan at $3/month (first month 50% off) with 30,000 TTS characters · Creator plan at $7/month (regularly $14/month) with 140,000 TTS characters · Pro plan at $70/month with 1,000,000 TTS characters
- API
- Not publicly listed
- Source links
- 16 recorded
Source-backed
Updated July 22, 2026Verified facts
Each fact points to a recorded source, making it easy to distinguish verified product information from claims that need checking.
Hume AI - Human Feedback for Voice, Speech, and Conversational AI | Hume AI
Human evaluation layer for voice, speech, and conversational AI · Multimodal emotional intelligence spanning 50+ languages, 48+ emotions, and 600+ voice des · Kairos Platform for evaluation suites, agent-to-agent and human-to-agent conversation simu
Human evaluation layer for voice, speech, and conversational AI
50+ languages supported
Multimodal emotional intelligence spanning 50+ languages, 48+ emotions, and 600+ voice descriptors
VoiceEQ - a benchmark for measuring the human quality of voice AI
Custom data collection built around specific scenarios, use cases, and evaluation requirements (Data Solutions / Build)
View 32 more verified facts
Kairos Platform - builds evaluation suites, simulates agent-to-agent and human-to-agent conversations, and tracks regressions over time
Expression Measurement API providing real-time emotion tagging and offline analysis across 48+ emotions and 50+ languages with 600+ output metrics
Human Feedback API providing pre-screened human raters returning per-sample scores, free-response feedback, and aggregated analysis in hours
Voice-native AI teams who need to know what callers are actually feeling
RW-Voice-EQ Bench with 4 leaderboards ranking voice AI models on Speech Recognition, Speech Understanding, Text-to-Speech, and Speech-to-Speech
SLM Judge Leaderboard ranking which small language models track human ratings most closely
Build and measure AI based on real human judgment, grounded in decades of research
Free plan at $0/month with 10,000 TTS characters and 5 minutes of EVI usage
Starter plan at $3/month (first month 50% off) with 30,000 TTS characters
Creator plan at $7/month (regularly $14/month) with 140,000 TTS characters
Pro plan at $70/month with 1,000,000 TTS characters
Scale plan at $200/month with 3,300,000 TTS characters
Business plan at $500/month with 10,000,000 TTS characters
Enterprise plan with custom pricing
Text-to-speech models Octave 1 and Octave 2
Speech-to-speech models EVI 3 and EVI 4 mini
SOC 2 Type II compliance
GDPR compliance
HIPAA compliance
Conversational AI
Expressive Text-to-Speech
Voice Conversion
Adding voice, expression, and empathy to applications
Engineers and developers
TypeScript
React
Python
Swift (iOS and macOS)
.NET
Open source SDKs, examples, and tools on GitHub
Hume API with documentation, tutorials, and API references
Hume AI
Practical capabilities
8 documented areasWhat it helps with
A concise view of the jobs, capabilities and integrations described in the recorded product sources.
Custom data collection built around specific scenarios, use cases, and evaluation requirem
Building evaluation suites, simulating agent-to-agent and human-to-agent conversations, an
Real-time emotion tagging and offline analysis across 48+ emotions and 50+ languages with
Obtaining pre-screened human rater scores, free-response feedback, and aggregated analysis
Generating expressive speech with natural emotion and customizable voices via Octave TTS
Transforming voices while preserving emotion via Voice Conversion
Adding voice, expression, and empathy to applications with a few lines of code
Benchmarking voice AI models via the RW-Voice-EQ Bench and SLM Judge Leaderboard
Availability
Source checkedWhere it runs and where to get it
Documented product formats, platforms and official distribution destinations. Availability can vary by region and plan.
Cost / license
Free plan at $0/month with 10,000 TTS characters and 5 minutes of EVI usage · Starter plan at $3/month (first month 50% off) with 30,000 TTS characters · Creator plan at $7/month (regularly $14/month) with 140,000 TTS characters · Pro plan at $70/month with 1,000,000 TTS charactersOpen source SDKs, examples, and tools on GitHub
Application types
Industry verticals served include AI companions, Automotive, Assistants, Interviews, PhoneExpression measurement models
Origin
The Hume Initiative is a not-for-profit developing concrete guidelines for the use of empa
Platforms
Kairos Platform - builds evaluation suites, simulates agent-to-agent and human-to-agent coTypeScriptReactPythonSwift (iOS and macOS).NETKairos, Simulate + Evaluate, Expression Measurement API, Human Feedback API, Data SolutionEmpathic Voice Interface (EVI).Octave Text-to-Speech.Kairos (Simulate + Evaluate), Expression Measurement API, Human Feedback API, Data Solutio
Adoption notes
DeploymentRelies on third-party vendors for hosting, cloud services, IT support, email software, and
LicenseOpen source SDKs, examples, and tools on GitHub
Model supportText-to-speech models Octave 1 and Octave 2 · Speech-to-speech models EVI 3 and EVI 4 mini · Octave · Provides speech models as part of its voice AI research infrastructure · Speech prosody model
Data controlHIPAA compliance · Uses reasonable organizational, technical and administrative measures; no system is guaran
Learning curveIntermediate
Primary use casesCustom data collection built around specific scenarios, use cases, and evaluation requirem, Building evaluation suites, simulating agent-to-agent and human-to-agent conversations, an, Real-time emotion tagging and offline analysis across 48+ emotions and 50+ languages with, Obtaining pre-screened human rater scores, free-response feedback, and aggregated analysis, Generating expressive speech with natural emotion and customizable voices via Octave TTS, Transforming voices while preserving emotion via Voice Conversion, Adding voice, expression, and empathy to applications with a few lines of code, Benchmarking voice AI models via the RW-Voice-EQ Bench and SLM Judge Leaderboard, Adding voice, expression, and empathy to applications, Bringing stories to life with human-like expression, empathic AI services, Emotion science research and empathic AI development, AI Companions — bringing digital characters to life with voices that feel human, from virt, Detecting voice fraud in real time., Preventing harmful deepfakes., Supports building foundation models, fine-tuning voice agents, and evaluating production s
What to verify before adopting
- Privacy policy does not apply to business API customers, research surveys/studies, or empl
- Research surveys and studies are covered by a separate Research Privacy Policy
- Privacy policy does not apply to employees or job applicants
- References the '30% problem' — fifty years of scientific focus on only six emotions covers
Evolution and major updates
4 dated updatesHume AI timeline
A concise history of software releases and material product changes. Events appear only when a dated source supports what changed.
Latest first · exact dates
Showing the newest updates and meaningful milestones. Open an entry for its summary and source.
Introducing Real World VoiceEQ benchmark
Open detailsHume introduced the Real World VoiceEQ benchmark for evaluating voice AI across ASR, TTS, speech-to-speech, and speech understanding, built from more than 1 million human ratings.
View source [17]Emotional Intelligence Is a Training-Time Property, Not a Prompt
Open detailsHume published an article arguing that emotional intelligence capabilities in voice AI must be acquired during model training rather than added through prompting.
View source [14]Building Voice Models Is No Longer a Modeling Problem
Open detailsHume published a product article explaining that voice model development now depends primarily on datasets, evaluations, and control rather than model architecture, and outlining its data and feedback offerings.
View source [13]What is semantic space theory?
Open detailsHume published an article describing semantic space theory (SST), the inductive, data-driven approach underlying its emotion understanding research and models.
View source [19]
Citation ledger
20 unique pagesRecorded sources
Facts, answers, structured details, milestones and primary resource links cite this shared ledger. Each external page appears once; release tags from the same GitHub project are grouped under one release history.
- 1hume.ai 16 facts · 4 answers · Official site
- 2hume.ai/developers 12 facts · 4 answers · Documentation
- 3hume.ai/privacy-policy 12 facts · 3 answers · Security
- 4hume.ai/blog/octave-tts-prompting-guide 12 facts · 2 answers · Official site
- 5hume.ai/blog/the-feelings-lab-ep-22-listener-questions-emotion-science-news 12 facts · 2 answers · Official site
- 6hume.ai/customers 12 facts · 2 answers · Official site
- 7hume.ai/about 12 facts · 1 answer · Official site
- 8hume.ai/pricing 12 facts · 1 answer · Pricing
- 9hume.ai/explore/speech-prosody-model 4 facts · Official site
- 10hume.ai/blog 1 answer · Official site
- 11hume.ai/api-data-usage-policy Documentation
- 12hume.ai/blog/announcing-evi-3-api Documentation
- 13hume.ai/blog/data-blog-jan 1 milestone
- 14hume.ai/blog/emotional-intelligence-is-a-training-time-property-not-a-prompt 1 milestone
- 15hume.ai/blog/hume-custom-model-api Documentation
- 16hume.ai/blog/introducing-hume-evi-api Documentation
- 17hume.ai/blog/introducing-real-world-voiceeq-measuring-the-human-quality-of-voice-ai 1 milestone
- 18hume.ai/blog/tutorials-hume-api Documentation
- 19hume.ai/blog/what-is-semantic-space-theory 1 milestone
- 20hume.ai/human-feedback-api Documentation
Research status120 substantive facts · 16 source pages · quality score 100/100