Skip to content
aroa.ai
about me
about
background and decisions
software
languages and tools
hardware
what i work on
projects
all of them
all twelve, on one page
data
pipelines and analysis
dev
tools i built for myself
writing
devlog
decisions and failures
blog
articles, soon
the model
model card
what it is and how it was trained
eval
where it works and where it fails
ES
Hi, I'm Aroa Index.
who are you?
what projects have you built?
why aren't you an LLM?
>
type
/
to see the commands
Aroa Index 1.7
▾
Aroa Index 1.7
Pure retrieval. Returns the corpus document that fits best.
512 KB
~10 ms
any browser
Aroa Serve
Same retrieval, but it writes on the server with the whole corpus in front of it. Combines documents and follows the thread.
0 KB
~1 s
needs a connection
Aroa Infer
Same thing, but a local LLM writes the answer with those documents in front of it.
~2 GB
slow first load
needs WebGPU
getting ready…
send