LAUNCH KEYNOTE
The model that sees, listens and answers in real time
Presented by {{author}}
01
Why real time matters
02
Meet Aurora
03
Under the hood
04
Building with the API
05
Pricing and access
PART ONE
WHY NOW
Today’s assistants pause, buffer and lose the thread. Aurora keeps up.
People don’t wait for each other to finish a paragraph. They nod, interrupt and change their minds halfway through a sentence.
Aurora listens while it speaks, so it can stop, adjust and carry on, just like the person across the table.
{{#}} / {{total}}
Understands screens, photos and live video as you share them.
Hears tone, accents and interruptions in over 40 languages.
Replies in a natural voice and acts through your own tools.
{{#}} / {{total}}
Stream voice and video in. Get speech, text and tool calls back.
import { Aurora } from 'aurora';
const aurora = new Aurora(env.AURORA_KEY);
const session = await aurora.live({
model: 'aurora-1',
senses: ['vision', 'voice'],
voice: 'ember',
});
session.on('reply', (r) => show(r.text));{{#}} / {{total}}
languages spoken and understood
faster than our previous model
uptime across every region
{{#}} / {{total}}
Aurora answers before the natural pause in a conversation ends.
TODAY
Private preview with 200 design partners
NEXT MONTH
Public beta opens to every developer
IN 3 MONTHS
General availability in 12 regions
WITHIN A YEAR
Aurora on device, with no network needed
{{#}} / {{total}}
{{#}} / {{total}}
Noor Haddad
Product lead, Tessellate Health
Mira Castell
Co-founder, CEO
Dev Rao
Head of research
Kofi Mensah
Voice and audio
Lena Strand
Platform lead
hello@aurora.example · aurora.example/docs