zackify

6 ms·

1.
by zackify 6d ago
I just have a slack bot running on a VM that sees a message and invokes pi. It would be trivial for every request to clone a full lxd container and have all the tools and repos required if I wanted to allow it to do even more. Not sure why
2.
by zackify 12d ago
What type of setup do you use? I have very small 4-5k initial context and do one task then clear. I rarely go above 150k context for most things.
3.
by zackify 12d ago
Yeah what, I'm using flash models and getting a ton of stuff done. 300-400b param size with pi.dev
4.
by zackify 14d ago
I have been using glm 5.3 flash and it feels as good as opus 5. Put a lot of work into it this week (100m tokens). Now I'm curious to try this one. These smaller models are getting very good imo
5.
by zackify 15d ago
It DOES still hurt. For example the eBay app. Does not allow installing from the play store on grapheneos.
6.
by zackify 16d ago
No memory. No web search 99% of the time. Two agent.MD files that are very small. One on each project. One at parent project level. Did 30M tokens through glm 5.3 flash today for 52c Using pi and a few extensions my initial context is alway
7.
by zackify 16d ago
I ran 30M tokens through for 50c... insanity that this is possible. and it really is opus 4.8 level.
8.
by zackify 17d ago
Any plans to offer a way to generate audio books on demand. like if i set kokoro, a book auto imports, it can make the file for me for the whole book after some time of local gpu work?
9.
by zackify 17d ago
Starlink? It's never gone anymore
10.
by zackify 17d ago
Wow I didn't even think about just using on device AI. I'm going to set this one up and try it with kokoro which has the most natural for small size that I've seen. Wonder if the paperwhite 12th Gen can handle it. And also se
11.
by zackify 17d ago
This is great. Recently got an xteink x3 and jailbroke my kindle, making custom stuff for it. I couldn't find it exactly in the docs. Sorry if I missed it. If you align these and then pull onto koreader via opds or other way, will audi
12.
by zackify 19d ago
You do, there's like 20 providers for any model on openrouter. You can also just spin bedrock or gcp and download the weights for later if you're worried. It's never going to make cost sense when the token rate is so low with
13.
by zackify 19d ago
https://voxtype.io/ https://tryvoiceink.com/ There's so many of these, at this point I've seen 10 clones make the front page each time as if there never existed local only options before. Also whi
14.
by zackify 21d ago
super sick, already setup a derp server. dns node key for fun on one and two machines that connect, with systemd to keep tailcat serving ssh running
15.
by zackify 21d ago
Can't even run self hosted github actions lol
16.
by zackify 24d ago
Recently jailbroke my kindle so I could have a camera pop up when frigate detects a person or a package while I'm reading. I think with omarchy adding easy to vibe code extensions and the way AI makes stuff so easy, I hope every OS giv
17.
by zackify 25d ago
I think the spec overcomplicates everything honestly. Its not that hard to add a long running auth token and put it in the MCP config as a header to send along and then avoid all the extra special rules. "Oh no it's a long lived t
18.
by zackify 1mo ago
The world doesn't make sense to me. I use 5.6 Luna. Deepseek v4 flash 0731. And kimi k3. Don't even need claude anymore at their insane prices for anything I do.
19.
by zackify 1mo ago
I have crossink which is awesome on the x3. I find it better with more features on top of crossink. And then setup home assistant and picture rendering so when a package or person is detected while I'm reading on a jailbroke kindle, it
20.
by zackify 2mo ago
I just use whatever's cheapest for personal things. Dsv4 flash and got 5.6 Luna are great for the price and I'll use the openrouter benchmark view to compare the overall score.
21.
by zackify 2mo ago
Yeah I use rcs and you can use the pixel camera app if you want?? Rcs had a bug that was fixed recently though
22.
by zackify 2mo ago
I have the llm twean every little piece of pi which is why I love it. And swapping models. Plus not having scroll jank like I got in opencode. I even tweaked hitting "enter" when nothing is in the input box, it resume the last req
23.
by zackify 2mo ago
Yeah wyoming protocol I think someone even made kokoro Wyoming. I forgot exactly what I did it'd been a little while
24.
by zackify 2mo ago
A Tesla can choose to auto lock. Or not lock. Or not lock in a certain area. Plus the app is an actual usable thing. Not a crap one that sends notifications about doors being unlocked in a garage that never updates. Last year I had to take
25.
by zackify 2mo ago
I use kokoro with home assistant and its great. I find its the most natural sounding and small too. I speak over sonos speakers when certain events happen. And use it as my voice assistant.
26.
by zackify 2mo ago
After seeing kia evs and having a Tesla. Its the only good EV brand because the software from everywhere else is a complete joke. Kia will tell me my doors are unlocked when I'm at home. Tesla has a set home feature. Plus the 50 other
27.
by zackify 2mo ago
$6 a month I plan to use deepseek v4 flash mainly which should provide closer to 5x the usage on the cheaper ones but no set number
28.
by zackify 2mo ago
I switched to yearly Cline pass because it was too cheap haha
29.
by zackify 2mo ago
Same. Btop is the best
30.
by zackify 3mo ago
I got mine! Did the reserve list 10 minutes in a few weeks ago. I like it way more than I thought.

More ›