zclaw: personal AI assistant in under 888 KB, running on an ESP32
23 points
by tosh
9 hours ago
| 5 comments
| github.com
| HN
v9v
7 minutes ago
[-]
[delayed]
reply
throwa356262
1 hour ago
[-]
"LLM backends: Anthropic, OpenAI, OpenRouter."

And here I was hoping that this was local inference :)

reply
micw
1 hour ago
[-]
Sure. Why purchase a H200 if you can go with an ESP32 ^^
reply
__tnm
27 minutes ago
[-]
haha well I got something ridiculous coming soon for zclaw that will kinda work on board.. will require the S3 variant tho, needs a little more memory. Training it later today.
reply
peterisza
51 minutes ago
[-]
right, 888 kB would be impossible for local inference

however, it is really not that impressive for just a client

reply
Dylan16807
36 minutes ago
[-]
It's not completely impossible, depending on what your expectations are. That language model that was built out of redstone in minecraft had... looks like 5 million parameters. And it could do mostly coherent sentences.
reply
theturtletalks
47 minutes ago
[-]
Is there a heartbeat alternative? I feel like this is the magic behind OpenClaw and what gives it the "self-driven" feel.
reply
g947o
1 hour ago
[-]
Serious question: why? What are the use cases and workflows?
reply
grzracz
17 minutes ago
[-]
I don't fully get it either. At least agents build stuff, claws just run around pretending to be alive?
reply
milar
1 hour ago
[-]
for fun!
reply
johnea
1 hour ago
[-]
I don't really need any assistance...
reply
throwa356262
1 hour ago
[-]
Me neither.

But I have 10-15 ESP32's just waiting for a useful project. Does HN have better suggestions?

reply
cameron_b
32 minutes ago
[-]
desk rover - https://www.huyvector.org/diy-cute-desk-robot-mo-chan

a kid-pleaser at the very least

reply
pacifika
1 hour ago
[-]
Build a synthesizer
reply
brcmthrowaway
1 hour ago
[-]
Why do you have so many? eWaste..
reply
iwontberude
35 minutes ago
[-]
No, no, but we insist!
reply