Active daily | Mutual minds #connect ๐ค๐ Local AI, systems, markets & ethics. Hardware over code, research till it works - serious tech with a playful twist ๐
@AnkanXplorer Exactly, itโs moving way too fast ๐
I love the pace of open source! but Iโm also waiting for someone else to load the latest beasts first so I donโt have to keep downloading 200+ GB models every week haha
New open-source competitor dropped today ,Hy3 looking strong ๐ฅ
295B MoE, Apache 2.0, rivals trillion-scale models on agentic/coding benches, free API trial running.
Did anyone try it on 2ร RTX PRO 6000 yet?
Or Iโll have to.. ๐
#LocalLLM#RTXPro6000
๐Hy3 is here.
295B MoE. Best in its size class. Rivals trillion-scale flagships.
Reliable and affordable for most agentic usecases.
Apache 2.0. Friendly for commercial use.
FREE API for 2 weeks โ https://t.co/EyURKwTdgi
๐ค https://t.co/twqJpqb2SL
๐ https://t.co/4uEkIU1cW4
@TheAIAuthority Yes, I planning to try it tomorrow. New strong models are dropping almost weekly now , I like that , but getting tough to decide which one is truly the best ๐
@AnkanXplorer Any hands-on trade like plumbing, welding, or HVAC. Robots and AI are great in controlled factories, but real world chaos, tight spaces, and unpredictable conditions still need humans
@jun_song Let us know how your tests go! Would love varied comparison points vs DeepSeek V4 Flash, especially on reasoning nuances and multi-turn reliability
Hi @X Algorithm ๐โจItโs Monday builders! Looking to #connect ๐ค
with founders/creators/enthusiasts in:
โข Entrepreneurship ๐ฅ
โข Startups ๐
โข Product Building ๐ ๏ธ
โข Web Dev ๐
โข AI & Automation ๐ค
โข Local LLMs open-source ๐ฉโ๐ป
โข Also helping small accounts mutually grow with good new friends. Drop what youโre interested in or working on ๐ @LocalAiCherry
Feeling playful this weekend, away from codes ai tech ๐
X turned mutuals into ride-or-die friendsโฆ
and unfollowers + non-mutuals into full-blown villains ๐๐ฅ
One day youโre laughing at each otherโs replies like youโve known them for yearsโฆ
but the second someone new hits unfollow? Suddenly youโre plotting their downfall like a movie antagonist ๐
This app didnโt just give us timelines - It gave us emotional warfare.
Tag a good mutual you know and letโs all #connect ๐ค
LongCat 2.0 is now open-source - 1.6T MoE beast!
Letโs go guys! ๐ Whoโs going to be first to test this?!
Or better question: who has that type of massive hardware?๐
#LocalLLM#LongCat
๐ฑ LongCat-2.0 is now fully open-source โ MIT licensed, no restrictions.
Since our launch a few days ago, the response from the community has been incredible.
Thank you for all the feedback, discussions, and interest.
Today, weโre releasing the model weights and inference code to everyone.
โ 1.6T MoE ยท ~48B active ยท 1M token context
โ Agent-native: Integrates directly with Claude Code, OpenClaw, and Hermes Agent
โ Deployment: Support both GPU and NPU platformsโ verified on large-scale domestic clusters
๐ Tech Blog: https://t.co/W9EYWDW3Y9
๐ค HuggingFace: https://t.co/EP2szwIhu2
๐ป GitHub: https://t.co/qJzXMQcC2z
๐ช ModelScope: https://t.co/nKt5Wh3m8Y
๐ Inference Code
GPU: https://t.co/Tq8QKvneH6
NPU: https://t.co/l96ebhFxN6