Michael🇺🇸

@ksuniri @HuggingModels Yes

@HuggingModels macbookpro m5pro 18cpu 20gpu 48gb 메모리에 충분할까요?

Ever wanted an AI that sees images AND talks without filters? Meet Qwen3.6-35B-A3B uncensored. This image-text-to-text model is abliterated, meaning it drops the usual guardrails for raw, uncensored conversations. It's a heretic in the best way possible.

@_nasch_ Eso corren en mi celular 🤣

Estoy corriendo qwen3.6 A3B en Q4 en un mini pc con gráfica integrada a 20 token por segundo. Lo dejé en mi red y ahora puedo usarlo desde cualquier computador local. El futuro de los LLM es ejecutarlo en local. 🤘🏻

@NousResearch Let give hermes jarvis powers ❤️

Even our toughest critics come around eventually

Event with @googlegemma next week! Hang out with the fellow builders and members of the Gemma + LM teams at LM Studio HQ in NYC. When: Monday, May 11th RSVP: required, link below High likelihood of pizza and cool demos

@gnukeith Worse expensive piece of crap

Best Chinese model right now? I'm using Kimi K2.6 myself

@mr_r0b0t Same thing with kimi

Be me: Sign up for DeepSeek API because everyone is saying how reasonably priced it is. Load up $50, let’s see. 30 minutes in, here’s where we’re at using V4-Pro

@DataChaz Europe mistral

USA has Claude USA has ChatGPT USA has Gemini USA has Grok China has Qwen China has DeepSeek China has Kimi China has MiniMax Europe India Seriously, what TF is the rest of the world even doing?!

@nv_sonti Thank you 🎉

@arismendius On 32GB you’re in the 7B–14B sweet spot. Qwen2.5 Coder is probably your best bet for BrowserOS.

@nv_sonti dude, what's the best LLM to use with BrowserOS? I have Ollama on my MacBook Pro with 32GB RAM, and I am having a hard time finding a suitable LLM. Thank you.

@quxiaoyin @AnthropicAI @nvidia And when you prompt the llm reply i am opus 4.7 🤣

RIP @AnthropicAI and @nvidia! DeepSeek v4 flash is $0.14/M input (with cache hit $0.0028) and $0.8 output/pro is $0.435 input and $0.87(cache hit $0.0036) versus Claude 4.7 $5 and $25. DeepSeek’s own official API, hosted with Chinese chips is 3x cheaper and much faster(!) than openrouter’s cheapest vendor. (Sources told me they use Chinese chips) Their playbook: make free open source model. Then make inference cheaper than hosting on your own GPUs. Anthropic is in big trouble. If you don’t believe me just try running your coding agents with DeepSeek APIs and pay $2 to test. It will last a few days.

#claude taking the day off.... Am i getting compensated for my time lost? #anthropic

@jtdavies @Prince_Canuma How much memory!

DeepSeek-V4-Flash on a Laptop. Everyone's doing it so I guess I have to too, I was curious. This is DeepSeek-V4-Flash (2bit-DQ) running at 43 toks/sec on my M5 Max Laptop using @Prince_Canuma's PR on Apple's mlx-lm.

@openclaw browseros.com

@deepseek_ai I asked your model, "Who am I talking to?" It responded #claude opus 3.0 🤣🤣🤣🤣

@satyanadella Make it work on mac

Agent Mode is here in Outlook! Copilot can now help run your inbox and calendar, triaging emails, rescheduling meetings, and helping you stay on top of what matters most.

@openclaw Make it stable and i will give it another shot

OpenClaw 2026.4.24 🦞 ☎️ Voice calls can now reach the full agent 🧠 DeepSeek V4 Flash + Pro join the team 🖱️ Browser automation got coordinate clicks + better recovery 🔧 Telegram, Slack, MCP, sessions, and TTS fixes More reach. Less duct tape. github.com/openclaw/openc…

@JulianGoldieSEO Expensive

🚨BREAKING: Hermes Agent v0.11 Just Changed AI Agents Forever! Hermes Agent just dropped V0.11. Biggest update ever. 🔥 → Brand new terminal interface. →Unlimited sub-agents spawning sub-agents. →GPT-5.5 now works inside it. →Access to every major AI model from one tool. →Redirect your agent MID-task with one command. This is a full rebuild. 🤯 FREE and open source ↓

@jun_song How much memory i need?

2개월 뒤에는 당신의 맥북에서 Opus-4.6 수준의 로컬AI를 구동할수 있을겁니다. 그것의 이름은 Qwen3.7입니다. 이 트윗을 저장하고 정확히 2달뒤에 보세요.

@freddier Hahaha reset stats everywhere

Claude Opus está respondiendo super veloz desde el anuncio de GPT 5.5... hmmmm ...

@CodeByNZ Cuz apple take their commission

why is claude on mobile so much more expensive . 😭😭😭