I asked Codex 5.4 to reverse engineer a DOS game with no source code. It’s been running for 6 hours, I can’t look away. It unpacked assets, disassembled the EXE, rebuilt the renderer, and built my childhood favorite SkyRoads in Rust! Now think of all the games we can revive. https://t.co/u8zidt0JlN

spent the entire day testing Qwopus (Claude 4.6 Opus distilled into Qwen 3.5 27B) on a single RTX 3090 through Claude Code. this is my new favourite to host locally. no jinja crashes. thinking mode works natively. 29-35 tok/s. 16.5 GB. the harness matches the distillation source and you can feel it. the model doesn't fight the agent. my flags: llama-server -m Qwopus-27B-Q4_K_M.gguf -ngl 99 -c 262144 -np 1 -fa on --cache-type-k q4_0 --cache-type-v q4_0 if you want raw speed, base Qwen 3.5 MoE still wins at 112 tok/s. but for autonomous coding where the model needs to think, wait for tool outputs, and selfcorrect without stalling, Qwopus on Claude Code is the cleanest setup i've found on this card. i want to see what everyone else is running. drop your GPU, model, harness, flags, and tok/s below. doesn't matter if it's a 3060 or a 4090, nvidia or amd. configs help everyone. let's push these cards to their ceilings. let's make this thread the reference.

AgentLens: A WebUI for local observability of agent traces 🔎 📝 Happy to share a side-project for detailed real-time inspection of what all your coding agents are up to. npx -y @roberttlange/agentlens --browser It locally monitors session traces and allows you to interact & terminate sessions in Codex, CC, OpenCode, Pi, Cursor Agents & Gemini CLI 🧑‍💻

Introducing CLI-Anything🚀 Making ALL software agent-native with one command. Today's software serves humans👨‍💻. Tomorrow's users will be agents🤖. CLI-Anything: bridging the gap between AI agents and the world's software. One command line to make any software agent-ready for OpenClaw, nanobot, Cursor, Claude Code, etc. GitHub: https://t.co/BlRymgR21a 🤔 Why CLI-Anything? CLI is the universal interface for both humans and AI agents: - Structured & Composable - Text commands match LLM format and chain for complex workflows - Lightweight & Universal - Minimal overhead, works across all systems without dependencies - Self-Describing - --help flags provide automatic documentation agents can discover - Proven Success - Claude Code runs thousands of real workflows through CLI daily - Agent-First Design - Structured JSON output eliminates parsing complexity - Deterministic & Reliable - Consistent results enable predictable agent behavior 💡 CLI-Anything's Vision: Building Agent-Native Software - 🌐 Universal Access - Every software becomes instantly agent-controllable through structured CLI. - 🔗 Seamless Integration - Agents control any application without APIs, GUI, rebuilding or complex wrappers. - 🚀 Future-Ready Ecosystem - Transform human-designed software into agent-native tools with one command. #CLIAnything #openclaw #nanobot #claudecode

mlx-vlm v0.4.0 is here 🚀 New models: • Moondream3 by @vikhyatk • Phi-4-reasoning-vision by @MSFTResearch • Phi4-multimodal-instruct by @MSFTResearch • Minicpm-o-2.5 (except tts) by @OpenBMB What's new: → Full weight finetuning + ORPO h/t @ActuallyIsaak → Tool calling in server → Thinking budget support → KV cache quantization for server → Fused SDPA attention optimization → Streaming & OpenAI-compatible endpoint improvements Fixes: • Gemma3n • Qwen3-VL • Qwen3.5-MoE • Qwen3-Omni h/t @ronaldseoh • Batch inference, and more. Big shoutout to 7 new contributors this release! 🙌 Get started today: > uv pip install -U mlx-vlm Leave us a star ⭐️ https://t.co/un61O8fEZd

I just spent the last 24 hours building... a free tool for you to create your very own handwriting font quickly in the browser (no logins, all local processing). This *entire* "SaaS" is a single-page HTML app - ready for you to try: ⁨https://t.co/p6t4n0kSbH⁩ Having tested it extensively with my own handwriting - including ligatures - I can definitely say that it works. ;) Download your own OTF, TTF, etc. when done. 100% created with @AnthropicAI Claude via web chat. Out of the 250+ free apps I've developed, this may have the most mass appeal.

@cb_doge @grok @Similarweb 🎯 https://t.co/zXD9tbQCTm

BREAKING: Grok recorded 13,840,000 app downloads last month 🔥 More and more people are switching to Grok. Try @Grok if you haven’t yet. 🚀 Download here https://t.co/Lj8Kpmxzhw https://t.co/gDT6FAV4Vx

Gasped today when I realised you can search for flights natively in the Gemini app https://t.co/AjgnCskzNP

Grok Imagine prompt: A generic robot with a Frankenstein-like appearance, featuring stitched skin, bolts in the neck, and a square head, sitting on a large boulder and crying with tears streaming down its metal face. Standing next to him is a kind fairy holding a magical... https://t.co/QXYA5Odcjn

🔺 Trump’s son buys $50 million worth of oil sector stocks: Two days later, his father started a war with Iran. Oil prices soared dramatically. 🔴 He bought before the bombs fell 🔴 He bought before Iran closed the Strait of Hormuz 🔴 He bought before 20% of global oil supply was choked 🔴 He bought before crude oil prices reached levels not seen since 2008… ⚠️ Is this a deal or a coincidence that turned a small investor into a billionaire…?

Le flux migratoire des experts dans les technologies s'est désormais inversé : il y en a plus qui viennent en Europe que partant aux USA. Il s'agit généralement là de dynamique de fond, extrêmement structurante à long terme. https://t.co/fEOCJxDnoV

am prepping our NVIDIA chat with @NaderLikeLadder and @KranenKyle and got excited for some incredible news coming for @NVIDIAGTC next week - would really tune in for the Jensen keynote on the future of **agent inference** https://t.co/s5zjx5G7Ns also got the @brevdev x NVIDIA story on the pod! dropping soon.

You can 'Imagine' anything ↓ Cinematic shots, design assets, fashion photography, concept art, 2D illustrations, 3D characters… Here's how: https://t.co/K9kcQMG905

Anthropic and OpenAI used woke ideology to program their models. It shouldn't be a surprise when their models are starting to become a threat to civilization. Their entire premise is to lie and cheat based on left-wing principles. https://t.co/ckcVxZH94e

13 months ago I ditched my G550 and replaced it with a @tesla CyberTruck. My only regret in that period… is not doing it sooner!!! https://t.co/YueUhdeB9X

The new Mistral 3 models in llama.cpp https://t.co/G4HoEGw85G

https://t.co/hMNFYm1QlE

We joined forces with NVIDIA to unlock high-speed AI inference on RTX AI PCs and DGX Spark using llama.cpp. The latest Ministral-3B models reach 385+ tok/s on @NVIDIA_AI_PC GeForce RTX 5090 systems. Blog: https://t.co/60yKKzNnoN

🚘 Best-selling vehicle in each US state (2025): AL: 🇺🇸 Ford F-Series AK: 🇺🇸 Ford F-Series AZ: 🇺🇸 Ford F-Series AR: 🇺🇸 Ford F-Series CA: 🇺🇸 Tesla Model Y CO: 🇺🇸 Ford F-Series CT: 🇯🇵 Toyota RAV4 DC: 🇯🇵 Toyota RAV4 DE: 🇺🇸 Ford F-Series FL: 🇺🇸 Ford F-Series GA: 🇺🇸 Ford F-Series HI: 🇯🇵 Toyota Tacoma ID: 🇺🇸 Ford F-Series IL: 🇯🇵 Honda CR-V IN: 🇺🇸 Chevrolet Silverado IA: 🇺🇸 Chevrolet Silverado KS: 🇺🇸 Ford F-Series KY: 🇺🇸 Chevrolet Silverado LA: 🇺🇸 Ford F-Series ME: 🇺🇸 Ford F-Series MD: 🇯🇵 Toyota RAV4 MA: 🇯🇵 Toyota RAV4 MI: 🇺🇸 Chevrolet Equinox MN: 🇺🇸 Chevrolet Silverado MS: 🇺🇸 Ford F-Series MO: 🇺🇸 Ford F-Series MT: 🇺🇸 Ford F-Series NE: 🇺🇸 Ford F-Series NV: 🇺🇸 Tesla Model Y NH: 🇺🇸 Ford F-Series NJ: 🇯🇵 Honda CR-V NM: 🇺🇸 Ford F-Series NY: 🇯🇵 Honda CR-V NC: 🇺🇸 Ford F-Series ND: 🇺🇸 Ford F-Series OH: 🇯🇵 Honda CR-V OK: 🇺🇸 Ford F-Series OR: 🇯🇵 Toyota RAV4 PA: 🇯🇵 Honda CR-V RI: 🇯🇵 Toyota RAV4 SC: 🇺🇸 Ford F-Series SD: 🇺🇸 Ford F-Series TN: 🇺🇸 Ford F-Series TX: 🇺🇸 Ford F-Series UT: 🇺🇸 Ford F-Series VT: 🇺🇸 Ford F-Series VA: 🇯🇵 Honda CR-V WA: 🇺🇸 Tesla Model Y WV: 🇺🇸 Chevrolet Silverado WI: 🇺🇸 Ford F-Series WY: 🇺🇸 Ford F-Series 🔎 @VisualCap via Edmunds (2025)

I hate this site The below blue checked account makes up completely fake stuff - that never happened - and it gets 2M+ views. No it didn't happen People treat as fact, thanks to having a $8/month blue check Fake news (often written by AI) spreads on this site like wildfire https://t.co/6rDMoDutT8

Physical AI is transforming smart logistics, and coordination is the new frontier. 🚚💨 Introducing RAPID by LG AI Research: Our new simulation platform that decouples layout design from agent behavior for ultimate flexibility. Cross-test AI models with ease! 🤖. #LG_AI_Research #EXAONE #LG #Physical_AI #AAAI_2026 Check out LG’s RAPID paper presented at AAAI 2026! 👉Blog : https://t.co/imHL8dgslX 👉Paper : https://t.co/0ZXTzG0QQR