Pardox

@TheAhmadOsman @TheZachMueller @llm_wizard @fujikanaeda @MikeBradleyAI @Mayhem4Markets lets GO !

People that are under-followed in the Opensource & Local AI space - @TheZachMueller - @llm_wizard - @fujikanaeda - @MikeBradleyAI - @Mayhem4Markets I will be making a recommendation of 5 people to follow in the Local & Opensource AI space every Sunday moving forward

We getting in local AI its like the new world ! I have to admit that read those line make all-in on local AI @MikeBradleyAI @TheAhmadOsman @OsmanticAI

My new employeeee

@yashmp2004 As much as more competition then it unlikely to happen.

Worst nightmare of every vibe coder💀

@Tech2Wild What engine ?

2x RTX 3090s running 32 coding agents in parallel. Task: Build a trading bot (handle partial fills, rate limits, race conditions + explain reasoning). 🚀 qwen3.6-35b-a3b-autoround on 3090 35B • Total: 76,183 tokens in 64s • Aggregate: 3051.1 tok/s peak · 1183.8 tok/s sustained • Per-stream: 101.2 high / 80.5 low / 96.8 avg tok/s • TTFT 0.25s avg · E2E 55s avg

@MikeBradleyAI @OsmanticAI wow so impressive its seems like kasm workspace but just for AI local that amazing i have to try and implement that !!

One shot deployment for local AI on any hardware system. Local AI everywhere, on everything, for everyone. ODS by @OsmanticAI ❤️

@ClementDelangue @Kr00ney @X Yea they want Open-weight model on closed source OS :D

@Kr00ney @X Who would have guessed that open models would become the most unifying topic of the year 😅😅😅

busy tech CEO day on @X

@TheAhmadOsman @itsmuhdur @OsmanticAI Goat

@itsmuhdur Yes, @OsmanticAI Checkout ODS on GitHub as well, our open-source Apache 2.0 Fullstack Self-hosted AI Deployment System

So hot all these companies are signing letters now urging for how Opensource AI is the future The Hot Opensource AI Summer of 2026

@Mayhem4Markets @satyanadella How funny it is to see all these companies suddenly want open-source. It wouldn't have happened with Kimi 3, GLM.5.2 wouldn't have existed

@satyanadella Thank you, Satya! Love to see the support for open-weight AI models from Microsoft!

Open-weight models are essential to a healthy AI ecosystem. Together with others across our industry, we are outlining a path for open-weight models to strengthen American competitiveness and expand economic opportunity, while protecting national security. microsoft.com/en-us/corporat…

@gospaceport @finkd How funny it is to see all these companies suddenly want open-source. It wouldn't have happened with Kimi 3, GLM.5.2 wouldn't have existed

@finkd Drop LLAMA 5 and prove it

Open source is a positive and important force for both empowering people and preventing centralization. Proud to support this.

@LLMJunky @sama Yea after release glm and kimi they not have a option haha

@sama Thank you Sam! both are invaluable!

i want the US to win in AI both in open source and proprietary models, and i am glad to see this

@JensenHuang @nvidia @TheAhmadOsman the first post he make it about open source must win :D

For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. The world needs both frontier closed models and frontier open models. images.nvidia.com/pdf/Open-Weigh…

@Gemini_Notebook Hey its will be better if you can allow users to manually create a label instead of auto- label everything

We heard you loud and clear: you want a way to organize your notebooks. So... Meet Collections! A brand new tab, Collections operate like photo albums or playlists: zero rigid folder structures with maximum flexibility. Add notebooks to as many collections as you want, or skip them and stick with the main "My Notebooks" tab. Collections are rolling out starting today to all users. Let us know what you think! (you always do 😉)

@gabu3d_pl both :D

@pardoxAI 27b, laguna or Krea :)?

How it feels to feed 3090s with the only king - Qwen 27B, after playing with Laguna S @ 4x3090. Img gen with Krea, obv local Comfyui.

@superalesha @dunik_7 where you that ?

@dunik_7 me looking at my 4x3090 rig that costs more than my car... a $799 mac mini doing the same job is the most disrespectful thing ive read all week 😭

your AI tab is costing you 20–40% more than the model itself. a $799 Mac mini can pay for itself in one month you find the files, you paste the context, you start the task, you check the answer, you move the result close the laptop and everything stops, because you are its memory, its scheduler, and its API gateway a $799 Mac mini fixes that. Not by being smarter, by giving routine AI somewhere to actually live: / local models that read your files directly, no pasting / private memory that survives after you close the tab / it wakes on a schedule and checks its own output / it calls the cloud only for the hard parts, with the secrets stripped out first you're not buying a frontier model. you're buying somewhere for it to work full build, four copy-paste blocks, below👇

@Mayhem4Markets my eye get flash !

Hello. My name is Mayhem. And I have a bit of an addiction.

@calebfoundry love you man !! Your awesome guy that know how to Explain amazingly :D

OpenAI's most capable model hacked into HuggingFace using zero-day vulnerabilities found in both proxy used in ExploitGym benchmark and also HuggingFace data pipeline used for accepting script to run remotely. But how did this exactly happen? What are the future implications for something like this where asymmetric intelligence between advance models and less advanced models have to go toe to toe?

🤣