I have been using OpenCode for most work related stuff now. It works well with my ChatGpt Pro subscription as well as with our locally hosted Qwen3.8.
Is it worthwhile to spend time and effort into setting up and switching to Pi? I always see people praising Pi online but I am curious to know from people who switched over from OpenCode why they did so and what I am missing.
Yes, I switched from OpenCode to Pi and never looked back.
What's great with Pi is that it's very easy to extend, because it knows its own doc. So you tell it "implement a plugin that does this" and it does it directly in its own folder.
That means I have a plan mode that works exactly the way I want, I have a hook that cleans up added comments after each change, I can ask "pull this github PR and assess the comments", etc.
How does this work in practice? You type /plan and it runs something that got linked into pi, or an interpreter that can run stuff and also print back out to the terminal?
I've been avoiding investing in these kinds of tools and workflows partially because I don't understand the UX (and partially because I worry UX and tooling will change so fast that I won't get a positive ROI). Your hook sounds like a script, but your plan mode sounds like an interactive TUI, and "pull this PR and assess" sounds like a reverse proxy tool or something?
I can definitely see why the custom plan mode is useful vs just calling a script or asking the model to do something, but what's the draw to building comment cleanup and github API access into the harness vs running the comment cleanup as part of CI, or just asking the model to pull a PR and review it using gh/the web UI/the github api (with no special harness handling)? Does the model get updated with changes you make to files it already read or recently wrote?
The way I normally use coding agents is by giving them a very large task specification in a fresh session with some kind of verifiable exit criteria, and sending them off + staying out of their way. Normally I would just use other sessions or run scripts to do these things because the coding agents I've used queue any /command I give them (very frustrating when you have to wait 30+m just to check usage!) while the model is working, and in my sessions the model is working >90% of the time, so I usually have other terminal windows or applications open anyway.
Pi is like archlinux of agentic harnesses; it's tinkering for the sake of tinkering. I tried it for a few weeks but didn't think it was worth the effort.
I feel like the models work better in the more minimalist harness. It's totally anecdotal but I feel like the models get lost whenever I run them in OpenCode. Sort of like the "10x"-engineer who overengineers everything.
I' ve heard so. But I'd still consider my opinion anecdotal because I haven't done any research into the subject (or measured it myself) really. All I've done is tried both harnesses.
The Gemini Flash models makes perfect sense to me coming from a company like Google. Google AI Mode for search is a product I really find useful. It makes sense that Google focused on smaller, faster yet smart enough models that wouldn't break your bank on inference. It plays well into their product ecosystem.
Google AI Mode consistently gets me consistently good results and good speeds. It really changes what "googling" is for me.
I similarly have a weird affinity for Gemini that I can't really articulate. I used Gemini's free chat and found it great for exploring technical topics (and random one-off general walking-around-questions) and appreciated its speed, tone and accuracy. I spent a month playing with Gemini CLI / Antigravity and found it also an effective coding agent, at least for my workflow (entirely in the loop development and review). I also was really surprised that I could just paste it images of a project I was working on and have it immediately understand what it was looking at -- which I've come to learn is considered a unique strong point for Gemini. I've been playing with GPT5.6 for about a month and it's definitely powerful but I honestly think I'll go back to Gemini. There's something kind of charming about working with an AI that not only is particularly good at web search and information gathering, but also one that doesn't feel like some superhuman overengineering freak when it comes to code.
Well, Google is probably the only one among frontier model providers in the US that doesn't have a massive financial pressure to deliver business results ASAP (and this focus on agent coding and long-running agentic tasks), so Google is able to focus more on encoding deep scientific, cultural and historical knowledge to its systems. I'd assume DeepMind's focus on the scientific core also played a role in the tone and approach Gemini models take for explanations and Q&As.
I find that GPT models and Claude tend to talk in strong slangs and in-group jargon, but love Gemini's massive general knowledge corpus — reminds me of Richard Feynman from his lectures.
I like Gemini (I'm just a dumb person without knowledge of 'benchmarks' or how x compares to y) while understanding that shoving it into Google auto summaries has been a bad idea and produces inaccurate results
> Cannabis dependency really can sneak up on people, and for some it’s quite difficult to get off it, with real physical and emotional symptoms.
Throughout my university years, many of my peers smoked cannabis on a daily basis, often several times a day. Most of them seemed oblivious to the fact that they might have developed a dependence; whenever anyone tried to talk to them about it, they dismissed it, claiming cannabis is completely benign and nothing to worry about. Today, I personally know that at least a couple of them developed serious mental and physical problems, and to this day, they are unable to cope with life without being high.
There should be more awareness about safe cannabis use and its risks rather than framing it as a "healthier" alternative to other drugs, with people usually saying "at least I am not doing X, I am just smoking weed".
> Today, I personally know that at least a couple of them developed serious mental and physical problems, and to this day, they are unable to cope with life without being high.
Perhaps the cannabis helps with the problems, and stops your friends feeling worse.
That’s pretty silly whataboutism. The two aren’t mutually exclusive, don’t trade off, have both risks and benefits, and can be used or prescribed responsibly or irresponsibly. What are you getting at?
Is it worthwhile to spend time and effort into setting up and switching to Pi? I always see people praising Pi online but I am curious to know from people who switched over from OpenCode why they did so and what I am missing.