brolecule!!

@0xPaulius bcz making it urself is better than using a slop generator

if ur taking 3-6 months to ship a good app ur doing something wrong. sorry

@lmstudiodevs overthinks

|￣￣￣￣￣￣￣￣￣￣￣￣￣ | | lms get qwen3.5-9b | |＿＿＿＿＿＿＿＿＿＿＿＿＿ | (\__/) || (•ㅅ•) || / づ

@timstarkdev @TukiFromKL Which qwen3.5?

@TukiFromKL On agentic terminal coding, Qwen3.5 reaches 52.5 while GPT-5.3 Codex hits 77.3 — that’s a big gap on the tasks that matter most for real-world engineering work. datacamp.com/blog/qwen3-5

Do you understand what just happened? Qwen dropped small models you can run on a $600 Mac Mini. Locally. No internet. No subscription. No company controls your access. Go do this right now: Download LM Studio Search Qwen 3.5 Grab the MLX versions Load them You now have unlimited AI on your own machine. Nobody can take it away from you. Not a company. Not a government. Not a terms of service update. Everyone's fighting over who controls AI. The answer just became: you do.

@ashen_one that is legacy qwen3 model

btw you can download this entire local model on your phone too works offline + only takes 1gb of space i have no idea how that’s possible app is called locally AI

do they have a camera is this a new prediction from qwen

@tolibear_ "sonnet 4.6 level" 9b .. sonnet level.. yea stfu

Holy shit. I'm now running a Sonnet 4.6-level LLM. Local. Private. No rate limits. 100% Free. That was way too easy.

@crackticker it's so tempting but every single one looks useful for the product i cant bring myself to imagine someone not knowing the price of something

fun fact, free e-ink display hack

@Crypto_Jargon this account is run by openclaw and an indian is the owner

OPENCLAW JUST DROPPED A MASSIVE UPDATE 🤯 OPUS 4.6 SUPPORT. MODEL FAILOVER. APPLE WATCH APP. 1M TOKEN CONTEXT. AND NOW… AGENTS CAN SPAWN AGENTS.

@0xSero What can I run with my 16gb mac mini usingthis

My goal for the year: make local AI easy and pleasant to use, on your phone, laptop, coding agents, discord, browser and even on ESP. You will be able to talk to an Apple watch, run a local model on call, get it coding for you, etc.. Kimi on 150gb vram GLM-5 on 150gb vram MiniMax-M2.5 on 48gb vram QuantForge lets you take any model, on any hardware. Select a target size, and calibration datasets and then prunes and quantizes it. Work on my macbook, I'm reaping and quantizing some tiny models. By the end of the year I will make it so me and anyone can get any model to fit any hardware. RN it uses local hardware but I will integrate with Prime Intellect. Going to add some features for sharing datasets, and building one out from many independent components.

@oliviscusAI i think the rumor about qwen feeding claude into its data is true.

China just dropped Claude Opus 4.5 level model that runs locally. And it’s 100% free & open-source.

@agenticgirl half of this doesnt work

Someone finally built a production-grade Agent OS in pure Rust. OpenFang isn't a Python wrapper or a generic multi-agent orchestrator. Here are the specs: → 137,728 lines of code. → 14 crates. → 1,767+ passing tests. → Zero clippy warnings. Because it's Rust, the performance metrics are insane. Cold starts happen in under 200ms. It uses just 40 MB of idle memory. It includes a SQLite memory backend with vector embeddings and canonical sessions. For me, this is what battle-tested infrastructure actually looks like. GitHub Repo is in the comments.

@elonmusk Shut up trumpie

Grok

@NACHOS2D_ @MartiniArt_ IF i go to the theater and see a movie made entirely out of AI, i am going to kms!!!

Seedance 2.0 generated this in minutes. Hollywood got cooked. Seedance 2.0 access @MartiniArt_

@JulianGoldieSEO openclaw tweeted and made the video

Everyone is overcomplicating AI agents. You do NOT need: – A Mac Mini – A $200/month cloud stack – Some insane hardware You need 8 minutes. OpenClaw lets you: • Run a 24/7 AI worker • Keep everything on your machine • Encrypt your data • Schedule automated tasks • Scale to multiple agents And it’s free. Most people will ignore this because it sounds “technical.” It isn’t. One line install. Few questions. You’re live. The people who understand this early will build leverage most can’t catch up to.

@VadimStrizheus waste of money

This small box is creating the next wave of generational wealth. This is your sign to start using OpenClaw. 🫡

@Amank1412 You can visibly see the cut every time

most AI video models stop at 10–15 seconds because generating clips is easy. Telling a story is hard. Someone just built an AI Creative Studio that pushes past that limit 2–3 minute cohesive stories, consistent characters, controlled pacing, and actual narrative structure.

@cgtwts "incredible at coding" propaganda pulled out of your ass. this codes as good as a 30b model.

Qwen 3.5 Medium models - Outperforms all previous Qwen models - Beats models that are 6× larger - smarter than Sonnet 4.5 - incredible at coding - Can run on almost any modern computer - completely free

@JorgeCastilloPr Dumbass

Neat progress indicator 🤯 unreal

@AlexFinn One of them is off

I have built the world's most powerful home AI lab My OpenClaw is now powered by: • 3x Mac Studio w/ 512gb memory • Nvidia DGX Spark w/ 128gb memory • Mac Mini M4 w/ 16gb memory I just added the DGX Spark to my Mac Studio cluster using EXO to handle prefill, dramatically speeding up AI speeds This will be running my OpenClaw swarm, which is currently 5 OpenClaws and 4 subagents. I plan on increasing this to at least 10 OpenClaws over the next 2 weeks This agent swarm will have 1 mission: be a 24/7 autonomous organization that produces value constantly. I will add more compute as necessary, including more Mac Studios when the M5 Ultra releases I do not plan on slowing down. This is the single most important moment in the history of this species, and I plan on capitalizing on it. My mission is to create a framework that enables everyone to experience abundance. Accelerate.

@heynavtoor "runs on an anything with a cpu" croaks on armv7

🚨 BREAKING: Someone just rebuilt the entire AI assistant stack in Zig. It's called NullClaw. The binary is 678 KB. It uses ~1 MB of RAM. It boots in under 2 milliseconds. No runtime. No VM. No framework. No garbage collector. Just raw Zig. Here's why this is absurd: → OpenClaw needs a $599 Mac Mini and 1 GB+ RAM → NanoBot needs 100 MB+ RAM and Python → PicoClaw needs 10 MB RAM and Go NullClaw runs on a $5 board with 1 MB of RAM. Same functionality. 0.1% of the resources. Here's what's packed into that 678 KB: → 22+ AI providers (OpenAI, Anthropic, Ollama, DeepSeek, Groq, etc.) → 13 chat channels (Telegram, Discord, Slack, WhatsApp, iMessage, IRC) → 18+ built-in tools → Hybrid vector + keyword memory search → Multi-layer sandboxing (Landlock, Firejail, Docker) → Hardware peripheral support (Arduino, Raspberry Pi, STM32) → MCP, subagents, streaming, voice, the full stack Here's the wildest part: Every subsystem is a vtable interface. Swap any provider, channel, tool, memory backend, or runtime with a config change. Zero code changes. It even encrypts your API keys with ChaCha20-Poly1305 by default. 2,738 tests. ~45,000 lines of Zig. Zero dependencies beyond libc. 100% Open Source. MIT License.