The first AI framework you can actually SEE thinking
Watch every thought, action and guardrail check light up in real time — then build your own agents on the same rails.
Every request pulses through the brain — see it yourself on the Living Brain →
Don't take our word for it — watch it think
A peek at the Agent Playground with “Show Thinking” enabled. Open it live and ask your own question.
Six ways to see inside the machine
Every feature opens a real page in this portal — no mockups, no dead ends.
One framework, every brain level
Flip the ELI8 toggle anywhere in the portal and the same system explains itself like you're 8 — or like you're the architect.
🧒 Kid mode (ELI8)
The robot's diary! It remembers what you talked about before — like remembering your favorite pizza place. 🍕
topic: memory system
🤓 Expert mode
SemanticMemory: a vector-backed knowledge store with cosine similarity retrieval, pluggable embedding providers.
topic: memory system
Try the real toggle in the top nav — it works on every page. 🧒↔🤓
Every system, one framework
Aetherin Core ships with the full stack — each piece has a live view in the portal.
From zero to agent in 30 seconds
Scaffold
One command generates a ready-to-run Aetherin project.
Run it
Hot-reload, playground and portal included out of the box.
Open it
Chat with your agent and watch it think in the trace view.
dev → stage → prod, one pipeline
The same config that runs locally deploys cleanly through CI to Vercel — Docker included.
dev
Local playground & hot-reload
stage
Preview deploys on every PR
prod
Vercel production + Docker