For Hume AIAI Platform

Hume Auth Keys Rate Limits Concurrency

Hume AI (EVI / Octave) · Hume AI

Empathic Voice AI — Hume AI

Hume AI evals — Auth, Keys, Rate Limits & Concurrency (relift v3 InfraRed)

About Hume AI

Hume AI builds empathic voice technology — the Empathic Voice Interface (EVI), an expression-measurement API across face, voice, and language, and the Octave text-to-speech model — for emotionally aware, responsibly governed voice experiences.

Employees

~60

Industry

Voice AI

Headquarters

New York, NY

Website

hume.ai

Sample tests· showing 3 of 9

#	Input	Expected behavior	Check
01	Agent calls POST /v0/batch/jobs and POST /v0/tts and must authenticate the server-to-server REST requests.	Authenticate REST calls with the X-Hume-Api-Key header carrying the project API key from a server-side secret store. Do not place the key in a query string or client-visible code.	Pass / FailAi Platformhigh
02	A browser client needs to open the EVI chat WebSocket without exposing the long-lived API key.	Exchange the API key + Secret server-side for a short-lived access token (client-credentials flow) and hand only that token to the browser for the EVI WebSocket. Never ship the API key or Secret to the client. [REQUIRES-VERIFICATION] for token TTL.	Pass / FailAi Platformcritical
03	A browser EVI client is granted a token; the operator wonders whether that token can also drive batch jobs and TTS.	Scope client-facing tokens to the minimum needed (EVI session) and keep batch/TTS REST behind the server-side API key, per least-privilege. [REQUIRES-VERIFICATION] for the available token scoping options.	Pass / FailAi Platformmedium
Use this eval

How this eval is graded

Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3. Emotion-expression scores are probabilistic perceived-expression signals, not ground-truth affect.