Pi 1.0 - Hackdot

Re:Pi 1.0 by whiteblossom on Thursday October 01, 2026 @03:58PM (#49926357)

It's been pretty useful for experimenting with local models due to its more light-weight approach. Though I respect the creators a ton, OpenCode has not seemed to work as well with models that run on consumer hardware

[ Parent | on HN ]

Re:Pi 1.0 by unsnap_biceps on Thursday October 01, 2026 @04:15PM (#49926541)

Which models are you finding the most success with?

[ Parent | on HN ]

Re:Pi 1.0 by whiteblossom on Thursday October 01, 2026 @04:48PM (#49926910)

qwen3.5 4b and 9b have both been surprisingly good for small tasks. a pattern I've been using is orchestrating pi agents with a script (that you can write using a frontier model) for common tasks. one script i use a lot is organizing photos from my photography shoots.

[ Parent | on HN ]

Re:Pi 1.0 by Asooka on Friday October 02, 2026 @02:32AM (#49930431)

What does a "script" mean in this context? Like a Python script that runs a series of agents with small tasks? I'm still getting used to doing things with AI, this is all very new to me.

[ Parent | on HN ]

Re:Pi 1.0 by hypercube33 on Thursday October 01, 2026 @05:43PM (#49927406)

I have more beefcake machines and Qwen 4.6 36B has been awesome for coding. Its fast for a local model, seems to get a lot right most of the time, just its slower than OpenAI/Claude/Cloud Hosted stuff since I dont have a 24GB+ GPU (I have Strix Halo and a 12GB GPU)

[ Parent | on HN ]

Re:Pi 1.0 by vhantz on Friday October 02, 2026 @01:46AM (#49930181)

How many token/s do you get with that setup?

[ Parent | on HN ]

Re:Pi 1.0 by hypercube33 on Sunday October 04, 2026 @04:58PM (#49957782)

60-70 tokens/sec Windows 11 LLM Studio 250 or whatever the models max is below that as context Qwen3.6 35B A3B 35B-A3B, gguf

[ Parent | on HN ]