tabbi

@Fvkepopstar @theBuoyantMan personally, i feel like i have been toyed with lately.. tasks/audits that claude has been seemingly dancing around finishing for weeks, deepseek accomplished in straight forward manner, in what seemed like moments. i am trying to build, not wait on session limits to reset.

@tabbislashcat @theBuoyantMan nooo way!! why not keep claude at $20 and ds at like $10/month? want real opinion

Tried Deepseek V4 pro side by side with Kimi K2.6 and GLM5.1 What costs around $6 on GLM costs around $3.5 on Kimi K2.6 and around $0.60 on DeepSeek V4 Pro. And the work done by DeepSeek V4 pro is like 90% of what GLM/Kimi accomplishes, and even after back and forth fixes - it ended up costing $0.6 overall. For a motivated engineer/developer/quant willing to work through the flaws, Deepseek is a godsend.

@Sentdex my paw has been wavering over the cancel button all day..

I do not miss claude at all

@MaxForAI this is the definition of flex. congrats though!! stoked to see this.

我觉得现在豪车、名表或者什么首饰的炫富方式已经过时了。 2026年的方式是：我在我家部署了一个满血GLM5.2🫡

@jun_song passionate meowpen-source /cat over here! 🐈

I want to connect with everyone who has passion on Open Source AI. Anyone who is building, or planning on future. Drop comments so I can invite you to the group chat. 💪

@mirandanover at this point, becoming an employee seems like the only way… any prompt we send is super limited without access to the actual models their developers are using.

i need to see how openai and anthropic employees prompt they know things we don’t even just for personal chat use

@dev_maims i think they are keeping the fun toys for themselves.

Seriously, how is Google still not leading the AI race despite having all the resources in the world?

came to @opencode GO to try glm-5.2 & stayed for the deepseek v4 pro. might be the /cat.. but i seem to be accomplishing (way)more than claude has been allowing lately. 🖤 @Zai_org @deepseek_ai

@jun_song deepseek has been my go to lately👌

People waiting for : > GPT-5.6 > Gemini-3.5-Pro > Opus-5 Me : > Deepseek-V4.1 > MiMo-V3 > Qwen-4.0 > Huawei personal hardware > M5 Mac Studio

@YashHustle_22 of course! no different than an architect not constructing their own blueprint.

Can you call yourself a founder if your entire product was built by Claude?

@art_zucker been itching to pounce off the claude fence.. imma take this as a sign.

For anyone still hesitating to leave claude code: Hermes agent is fucking lit with glm5.2🔥

iKitty messages of encouragement from my love.. now i’m thinking i need to build a catbot for in-house deliveries. 🖤🐈‍⬛ #ai #dev #queertech

@araseb_ honestly, it’s simply obsession with creation.. i’m designing to entirely customize my life. if the rest of world digs on it eventually, that’d be cool, too.

be honest. are you building because you enjoy it or just for the money?

@remilouf 🧐 … you did what just crossed my mind. thanks for the sign!

Not paranoid but GLM 5.2 downloaded and stored ✅

hm.. everyone’s meowing about glm5.2 ✨ this /cat is curious @Zai_org 🐈

this past week has been the wildest ride of service being provided. my sonnet 4.6 couldn’t decide if it had 200k, 500k or 1m context windows. each time i signed in, it was a different size. i thought my brain is breaking… @claudeai

@SavinoDrip @rileybrown this literally crossed my mind as i started reading through the responses.. it’s decided. sell the house!

@rileybrown Should I sell my home for local compute? The new American dream is compute it seems.

If I had a budget of $10,000 to build a local setup to run GLM 5.2 locally AND any future local model that comes out soon what should my computer set up be?

@elijs_dev it worked.. this kitty is obsessed with all of the above.

Querido algoritmo please haz lo tuyo y conectame con otros locos obsesionados con programación, IA y startups

@nutlope Now.. what if, you throw the /cat at it? Kitty be curious. 🐈‍⬛

This model is insane at design. I asked GLM 5.2 (left) and Opus 4.8 (right) to build me a landing page and you can't even tell the difference. GLM cost $0.06 while opus cost $0.49. More than 6x cheaper while being faster + more token efficient. Another win for open source AI.

while everyone is talking; the slash cat is slinking. pouncing into the future of connection, soon. #ai #dev #comingsoon #queertech