Simple pricing for classrooms, schools, and museums
Free to start. $129 a year for your classroom. School plans from $600.
- No credit card to start
- Students always join free — no accounts
- Cancel anytime
- Purchase orders and check payments accepted
For trying Humy with one class
Explore
See what happens when your students can question Lincoln himself.
Includes:
- A starter library of AI historical figures, from Cleopatra to Abraham Lincoln to Martin Luther King Jr.
- 20 AI messages a month — enough to interview a figure together on the projector
- Auto-graded assignments in 5 formats: interviews, debates, simulations, essays, and short questions
- Panels: up to 10 figures in one chat, moderated by your students
- The Community library: figures, sources, assignments, and collections other teachers share
- 90+ AI tools, including Lesson Planner, Quiz Generator, and Reading Level Adjuster
- Sign in with your Google or Microsoft account
Limits:
- 1 teacher account (students are always free)
- 3 custom figures, 3 assignments, 3 collections
For your classroom, all year
Plus
The full figure library, voice calls, and enough messages for the whole year.
About $11 a month, billed yearly
Everything in Explore, plus:
- All 1,200+ AI figures, including scientists, authors, and native speakers for classes beyond social studies
- A yearly pool of 12,480 AI messages for your classes — 52 times the free plan
- Live voice calls with historical figures (calls draw from your message pool)
- Full transcripts of every student conversation, exportable to PDF or DOCX
- Standards-aligned assignments: AP, Common Core, IB, TEKS, and 8 more, or upload your own rubrics
- Up to 50 custom figures built from documents you upload
Growing team?
- Add colleagues anytime at the same per-teacher price (up to 15 seats self-serve)
Includes a 30-minute onboarding call with a Humy specialist.
For whole-school and public rollouts
Schools & Organizations
Roll out AI-powered learning across your school, district, university, or museum.
Every quote comes with a pilot plan
Request a quote- Everything in Plus for every teacher, plus volume discounts and a dedicated account manager
- Student-safe by design: FERPA- and COPPA-aligned DPAs signed in 16 states, and student data never used to train AI or for ads
- IT approval packet, DPAs, and vendor paperwork ready to forward — most schools are up and running within days
- Teachers sign in with Google or Microsoft (school SSO on request); students never need accounts
- Usage reporting and conversation oversight across your organization
- Custom figures built from your own archives, with quotations checked word-for-word against the original documents
- Deploy on the web or via QR codes, kiosks, embeds, and the API — in 50+ languages
We built the AI Hatshepsut experience for South Florida PBS. Ask about museum pilots.
“The world religion collection blew us away. It’s worth every penny, and I’d fund it myself!”
Laura Odom
English Teacher
“Humy saves me 4 hours per week on grading, and the feedback students receive is awesome.”
Frank B.
8th Grade History Teacher, California
“Unprecedented engagement! My students were so involved, they forgot about gym class, which never happens.”
Corey Pendleton
Headmaster
AI models
9 AI models, and you choose
Humy runs on the latest models from OpenAI, Google and Anthropic. For every figure you share, pick which models students can use and which one they start on. Put a light model in front of a big class and keep a deeper one for research.
- OpenAI
- Anthropic
Everyday
Fast and light on your plan. Good for a whole class asking quick questions.
- GPT-6 Luna
- Last Exam score
- 20%
- Gemini 3.5 Flash Lite
- Last Exam score
- 19%
- GPT-OSS 120BOpen-weight
- Last Exam score
- 20%
Advanced reasoning
Better use of sources and longer, more careful answers. Uses more of your plan per message.
- Gemini 3.8 Flash
- Last Exam score
- 48%
- GPT-6 Sol
- Last Exam score
- 46%
- Claude Sonnet 5
- Last Exam score
- 30%
Deepest reasoning
Highest quality for research projects and hard questions. Uses the most per message.
- Gemini 3.1 Pro
- Last Exam score
- 47%
- Claude Opus 5.5
- Last Exam score
- 56%
- GPT-6 Astra
- Last Exam score
- 53%
Every model draws on the same message allowance in your plan. Lighter models use less of it per message, deeper ones more, so the choice is yours for each figure.
Last Exam score is the share of Humanity's Last Exam a model answers correctly without tools: about 2,500 expert-written questions across 100+ subjects, history included, built so the answers can't just be looked up. Artificial Analysis (Independent benchmark results)
Also inside Humy: Gemini 3.8 Live for real-time voice calls; Gemini 3.1 Pro for page-by-page transcription of scans, including handwriting. Students always see which model they're talking to before their first message, and none of them train on student data.
Used by 90 schools, districts, and universities
