REAA Technologies
45 posts

REAA Technologies
@reaatech
Production-grade AI for small business. Free, open-source, no vendor lock-in.
Las Vegas, Nevada เข้าร่วม Haziran 2025
263 กำลังติดตาม5 ผู้ติดตาม

@araseb_ Actual stupidity? 1984 "ignore your own eyes and ears" for AI from our Greatest Ally.
English

@ItsDymant @TTrimoreau Nope, I hate Anthropic and so do the people of Gaza
English

@adahstwt The Chinese have already won! 😂 DeepSeek, Qwen, Kimi, GLM, MiniMax, etc. All are better and cheaper, and they don't defend the actions of our Greatest Ally.
English

@DavidOndrej1 Don't use that either. 😂 Kimi-K2.7, DeepSeek-V4-Pro, GLM-5.2, MiniMax M3, etc. All are better and cheaper.
English

@stelzner_n1150 I'm going with positive, because it showed everyone the true bigoted demonic colors or our Greatest Ally.
English

@henokcrypto It amazes me how so many people still have yet to try Kimi-K2.7 and it's TRILLION context and 6x Fable's speed. The pros are not waiting for Fable, the problem has already been solved.
English

glm 5.2 might surprise a lot of people and me included
5.1 was already sitting at the top of swe-bench pro for open models, and now 5.2 adds 1m context + max reasoning is a combo that matters in coding and it is where long context + stronger reasoning compounds..
its not just a bigger window but its holding the whole repo, the intent, the bugs, the tests, and the plan in one pass..
I expect it to land above gpt 5.4 / opus 4.7 on coding, and possibly outperform opus 4.8 and gpt 5.5 in specific agentic workflows...
earlier i posted a low deepswe score prediction for and the math said ~25, so i called ~25. ..but that logic only measures single-pass swe tasks and it doesn't capture what the 1m window actually unlocks..
deepswe scores single-pass tasks and its too short to ever fill the context, so 256k and 1m look identical on it...
but real agentic work fills the window, and that's where the model either holds the whole thread or starts losing it... that's the part the benchmark can't see and that's why i'm revising up..
we will know this week where glm 5.2 stands, im very optimistic about this one

Z.ai@Zai_org
Intelligence should be open, accessible, and ready to build with, empowering every developer, everywhere. GLM-5.2 is now available to all GLM Coding Plan users, including Lite, Pro, Max, and Team plans. docs.z.ai/devpack/latest… As our new flagship model, GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks. API and Chatbot services will launch next week. The model will also be officially open-sourced next week under the MIT License. The future of AI is open, and it belongs to the people.
English

@gmoneyNFT Sound slike you haven't tried Kimi-K2.7 with it's TRILLION context - release Friday after Fable died.
English

@KaiXCreator They're just trying to stay relevant, all while knowing they already lost to the Chinese.
English

@sflorimm Gemini for what, slop code? 😂 Use Kimi, DeepSeek, GLM, MiniMax, etc.. All are better and cheaper.
English

@Nikkon2025 @brockpierson Until they collapse the middle class to reset and send us all back to the Stone Age for those of us slaves that survive the genecide. Read World Economic Forem Agenda 2030
English

@syhily Kimi is awesome, glad to see someone else talking about it. GLM is still too damn slow.
English

@TiffanyFong He deserves it! He built and threw every rocket into space with his own strength. No help from anybody else! 🤣
English





















