Sabitlenmiş Tweet
I've been enjoying Milogs.
This one I'm very proud of.
This has the potential to make you a 10x human, especially if you have ADHD.
I barely have scattered brains and foggy days since I started using it fully.
English
Osahenrumwen Aigbogun
224 posts
Osahenrumwen Aigbogun
@leonard_osahon
Software Engineer | Building https://t.co/S06ZylWFNr @pillardash
Lagos, Nigeria Katılım Temmuz 2023
27 Takip Edilen36 Takipçiler
I can tell you from experience too, Kimi 2.6 is sufficient. Going back to codex after being on opencode/Kimi for a while made me realise that codex adds to what makes gpt models better, the harness. Specifically the desktop app; if I can replicate that on opencode somehow, I'll probably barely use codex anymore
English
Osahenrumwen Aigbogun retweetledi
Cancelled both my Claude Code Pro and ChatGPT Pro for this.
Kimi K2.6 is just as good for my side projects as Opus or GPT 5.4 were. The price for this is crazy low, and there are a bunch of models I can try (like DeepSeek).
Bonus: I'm moving away from building everything on Claude Code - now that both @opencode and @cursor_ai have their SDKs open, I feel I can rebuild the agentic workflows I built for Claude Code in a more platform-independent manner.
Lotto@LottoLabs
Update on Opencode Go It’s great value for $5/month, there’s really no reason not to do the first month. At $10/month it’s still good value and gets you access to all sota OS models. You can’t daily drive it without hitting limits on the big models but w/ Kimi x3 you won’t hit limits unless you’re insane. Overall highly recommend the first month, then make your own decision.
English
@nb4ld How are you getting good reliable output and daily use from these Q4 models? I've tried running it via LM studio on opencode, but tool calling fails 9/10 times. Any ideas?
English
A single V100 from 2017, half the price of a 3090, runs qwen 27b dense q4 with 256k context at double the speed of a 3090
Sudo su@sudoingX
most of you don't know how big a deal it is that a single rtx 3090 from 2020 runs qwen 27b dense q4 with 256k context at 40 tok/s, full agentic loops on hermes agent, zero tool call failures. the more i build on this card the more i think nobody really knows how untapped it actually is. the silicon was always capable, the models finally caught up.
English
Osahenrumwen Aigbogun retweetledi
@opencode My only problem is that I use kimi 2.6 too much and my tokens don’t last
English
One of the best feelings in the world comes when you are able to reproduce a bug that you discovered by mistake. It's even sweeter when you are able to reproduce it across various environments or platforms
English
Osahenrumwen Aigbogun retweetledi
Update on Opencode Go
It’s great value for $5/month, there’s really no reason not to do the first month.
At $10/month it’s still good value and gets you access to all sota OS models. You can’t daily drive it without hitting limits on the big models but w/ Kimi x3 you won’t hit limits unless you’re insane.
Overall highly recommend the first month, then make your own decision.
English
@enunomaduro Just use Kimi 2.6. Don't stress over anything else
English
Osahenrumwen Aigbogun retweetledi
@StefanTMD At times I wonder if opencode's $10 plan is profitable. It feels like I'm cheating or something
English
opencode with the $10 plan is a really good deal
Sam Altman@sama
codex with the $20 plan is a really good deal
English
This is honestly the best way to go.
I tried this setup for a while and I tested with Gemma 4 E4B - Q4_K_M variant on LMStudio using Pi as my harness. I wasn't all that impressed with the result.
I guess I'll try again with a higher spec device and model variant for best result.
Thanks. Saved for reference!
English
Lately I've been having fun with running coding agents fully locally. The setup I landed on is:
- Pi agent
- Gemma 4 26B A4B
- Server of choice: LM Studio/Ollama/llama.cpp
I wrote a step-by-step guide with instructions on how to set it up: patloeber.com/gemma-4-pi-age…
English
@opencode Nice! Thanks!
I see why they named themselves moonshot!
Kimi to the moon!
English
Okay it's official, Kimi K2.6 3x usage on Go for another week
OpenCode@opencode
hmm should we keep kimi k2.6 at 3x usage for another week or nah?
English
hmm should we keep kimi k2.6 at 3x usage for another week or nah?
OpenCode@opencode
it's settled.. 3x in go for 1 week
English
The more I use kimi k2.6 the more I want to find a way to download it and keep it secretly before it is taken away from the internet.
It is just wild how good the model is, and the fact that it's opened is simply amazing.
Thank you @Kimi_Moonshot and @opencode
English
I needed to update this tweet.
The more I use it, the more I realise it leaves Gemini 3.1 pro preview in the dust
Osahenrumwen Aigbogun@leonard_osahon
kimi k2.6 is better than gemini 3 flash. quote me
English
Osahenrumwen Aigbogun retweetledi
the more i use @opencode go the more i’m convinced this sub isn’t profitable for them they’re offering it out of love for devs
Jay@jayair
DeepSeek coming to Go in 3, 2, 1…
English
Was working on a feature using minimax 2.5 on opencode and only noticed when I gave it a screenshot; all the while I thought it was gemini 3 flash.
That's actually impressive how far these open-weight models have gone.
God bless you all working on it
English
WebKit 🤦🏾♂️ is your intention to drain my life force? How can a simple CSS rule work well on chrome, Firefox, SAFARI WEB but fail to work on WebKit webview!! Make it make sense!
English
