Zclaw – The 888 KiB Assistant
73 points by kristianpaul 3 days ago | 40 comments

tehsauce 7 hours ago
“888 KiB Assistant” but the assistant itself is a multi terabyte rental-only model stored in some mysterious data center.
reply
seertaak 2 minutes ago
The whole point is that this fits on an ESP32, which has wifi. We're not quite at the point where it makes sense to run the whole thing locally - if you do try it, it will need a fan, and be loud etc.

For my part, I installed Nanoclaw on my Arch derived OS (I love Arch!), and it worked fine until the next day some update decided to revert the power management settings, and now my glorious assistant is dead.

There's something to be said for a barebones OS. No bullshit, no updates.

Also, playing with hardware watchdog timers and GPIOs and DACs can be so much fun.

reply
amelius 7 hours ago
I'm getting "serverless" flashbacks.
reply
Rebelgecko 6 hours ago
It seems to support connecting to your own LLM on the same LAN
reply
croes 4 hours ago
The point is the agent is still the LLM. No LLM, no agent.
reply
dheera 4 hours ago
I tried connecting OpenClaw to ollama with a V100 running qwen3.5:35b but it was really, really, really slow (despite ollama itself feeling fairly fast).

These "claw" agents really multiply the tokens used by an obscenely huge factor for the same request.

reply
jcgrillo 2 hours ago
i recently decided to get into this ocean boiling game too, the 32GB V100 seems like a pretty good VRAM/$. if i may ask, do you make any special accommodations for cooling? i've never dealt with a passively cooled card before and i'm curious whether my workstation fans (HP Z840) will be sufficient. i'm going to try 2 cards at first but i think i might be able to squeeze a third in there
reply
dheera 34 minutes ago
No. I have a Titan V CEO edition, which is basically a 32GB V100 but has full active fan cooling which I'm finding works just fine.
reply
jcgrillo 13 minutes ago
Oh very cool. Some folks are printing shrouds for dual 40mm fans so I'll probably try that if the stock case fans don't do it
reply
kristianpaul 6 hours ago
My model is at home... just 16Gb still a lot but just FYI
reply
0xbadcafebee 5 hours ago
For people who don't get this: it's a Home Assistant type thing. You don't do inference on it, you send it a message on Telegram and it does things with physical things through GPIO. You could use a $140 Raspberry Pi with 8GB RAM and host a local model on it plugged into 30W AC power... or you could use a $10 ESP32 which can run for weeks on a tiny battery, and your existing Wifi connection with a cheap cloud model (cloud models are as cheap as $0.02/1MTokens). This makes it easier to ramp up on new ESP32 projects. You can just tell it to do things / give you info, rather than having to write code.
reply
newswasboring 3 hours ago
The model can be taken care of in cloud but hardware also depends upon what we want to do right? If we want to run some lightweight python scripts etc, we cant use ESP32 right?
reply
jcgrillo 2 hours ago
dope so i can hook it up to my appliances, hvac, etc? maybe it'll do rad things with my smart doorlocks, remote car starter. can it send my self-driving waymobile to pick up my family at the airport too? such wow
reply
amelius 7 hours ago
Me: "GPIO 5 can be active for a maximum of 100ms, then it needs to cool down for at least 1s. Otherwise the MOSFET is fried."

Zclaw: "GPIO 5 is active now, however the server is not responding so I'm awaiting further instructions."

reply
gas9S9zw3P9c 7 hours ago
I fail to understand why 888 KiB matters if it's just a wrapper around a cloud api.
reply
ramon156 7 hours ago
Have you seen OpenClaw's codebase? 680.000 LOC.

I care how big it is.

reply
Rebelgecko 6 hours ago
A lot of the *claws emphasize binary size and lines of code. I think for better or worse people treat codebase size as a proxy for "how much of the project is unsupervised, unmaintainable, buggy AI slop?"
reply
boznz 3 hours ago
8 is lucky number in China
reply
mihaelm 7 hours ago
Because of resource-constrained environments, the primary deployment target seem to be microcontrollers. You can get ESP32 boards for pretty cheap.
reply
renewiltord 7 hours ago
Because it means you can run it on an ESP32 which is a low power microprocessor package.
reply
Retr0id 5 hours ago
888KiB is quite large, but I see they're including the whole rest of the firmware in that number, fair enough. Their actual application code weighs only 35,742 bytes, compiled.
reply
hidelooktropic 7 hours ago
There are many concerns and areas for improvement with open claw and other similar projects (continuous loop script with broad OS access that manages your agents and interfaces with a standard messaging app)

However, file size I have never seen on that list. I would rather offer for something that is even bigger in file size so it afford certain functionality like better security tighter permissions however it would do that.

reply
stavros 6 hours ago
I made a secure one:

https://github.com/skorokithakis/stavrobot

Everything runs in containers (I run it on a server along with everything else), plugins have a permission system so eg the AI can read emails but not delete or send, etc.

I really like it, I run it as my main agent and it has been extremely helpful.

reply
croes 4 hours ago
Part of the usefulness is based on the same thing that makes it so dangerous.

If it can only read but not act, it’s safer but less useful.

reply
stavros 4 hours ago
I can't restrict OpenClaw if I don't need the extra capabilities. I can restrict this.
reply
mihaelm 6 hours ago
File size is a legit property to keep in mind if your goal is to create an agent that runs on ESP32 boards. They don't expect you to run Zclaw on Mac Mini.
reply
allthetime 6 hours ago
What's the use case for running this on a tiny board? Isn't the whole point that it can use your computer for you?
reply
mihaelm 6 hours ago
For something like OpenClaw yes, but not for Zclaw. I think the naming is more about riding the current wave of Claw-related interest rather than positioning it as competition or replacement for other clawies.

Zclaw is about running an agent in your embedded system.

reply
markstos 6 hours ago
The examples seem to suggest it would be chatting with your home automation in natural language.
reply
fc417fc802 5 hours ago
Before you know it your smart thermostat will be blogging. The joke is on everyone who thought IoT couldn't get any worse. Just imagine the new landscape of security vulnerabilities this opens up.
reply
markstos 5 hours ago
My "smart" gas stove can be turned on over the internet (if I allow it to connect)—perfect appliance to put an LLM in charge of.
reply
jesse_dot_id 7 hours ago
Clicked on this expecting to see a crontab file.
reply
wartywhoa23 6 hours ago
Still lightyears from a one-bit AI assistant. Send 1 to save the humankind, 0 to exterminate. And hurry up because it's in undefined state right now!
reply
cwoodyard 5 hours ago
[dead]
reply
giancarlostoro 7 hours ago
The domain crashed and burned or something, hopefully this link is correct:

https://github.com/tnm/zclaw

reply
jFriedensreich 8 hours ago
There is the same divide starting to form that NFTs had back in the day. Tech bros instantly like if something has claw in the name, the rest of us will dismiss anything with that naming and philosophy as toxic slop culture. will be interesting to see how far this one will go.
reply
4ndrewl 7 hours ago
Is Clawcoin a thing yet?
reply
karolcodes 7 hours ago
1. why
reply
dpe82 6 hours ago
2. fun
reply
rvz 6 hours ago
2. hype
reply