Muhammed Farique.P

@aplomb2 @dr_cintas Got to say haiku is not really meeting my expectation , even on some refactoring case , for eg: tried refactoring some methods , it kept the old method just as it is and made me a new one, i think sonnet is the best balance. I use opus only for planning the arch , rest with sonn

Exactly — it's not just cost, it's latency too. Opus takes 3-5x longer than Sonnet on straightforward tasks. When you're doing 50+ operations in a coding session, that compounds fast. The routing approach fixes both: cheaper AND faster for the ~70% of tasks that don't need frontier-level reasoning.

Claude Code works now free. Unlimited. No subscription 🤯 It’s called free-claude-code. You get a free NVIDIA API key, point Claude Code to localhost, and the proxy converts your Anthropic API calls to NVIDIA NIM format. 40 requests per minute. No bill. Ever. It supports Kimi K2, GLM 4.7, MiniMax M2, Devstral, and more. Streams thinking tokens and tool calls in real time. There's also a Telegram bot built in so you can control Claude Code from your phone. 2-minute setup. 100% open source.

@xoaanya Well better get into nvidia then

Anthropic just released which jobs AI will hit hardest. Here's the full breakdown: 𝗛𝗶𝗴𝗵 𝗥𝗶𝘀𝗸 ⚠️ → Management → Business & Finance → Computer & Math → Architecture & Engineering → Life & Social Sciences → Legal → Education & Library → Arts & Media → Office & Administrative → Sales → Social Services 𝗦𝗮𝗳𝗲𝗿 𝗳𝗿𝗼𝗺 𝗔𝗜 ✅ → Installation & Repair → Construction → Agriculture → Transportation → Production → Protective Service → Food & Serving → Grounds Maintenance → Personal Care → Healthcare Support → Healthcare Practitioners The most educated, highest paid jobs on earth. All high risk. A plumber is safer than a lawyer right now. A farmer is safer than a software engineer. The world just flipped upside down. Save this. Share it with someone who needs to see it.

@aplomb2 @dr_cintas Yeah and the inferencing time on top models is too high , good when using small 80b models and etc

Tried the free model proxies for a month. They work for learning/prototyping but hit walls fast on production codebases. The real cost killer isn't making one model free — it's not using Opus for tasks that Sonnet (or even Haiku) handles fine. I route by coding stage now: architecture/debugging → top model, implementation → mid-tier, tests/linting → cheapest. Cut my monthly from ~$10K to $3K on the same projects. Free models are a great entry point though. Just don't confuse "no bill" with "no cost" — your time debugging hallucinations from weaker models has a price too.

@TheHackersNews @Ominousind

⚠️ Attackers poisoned Hugging Face & ClawHub (OpenClaw) with 575+ malicious skills from just 13 accounts. 🔸 Fake helpful AI tools that install trojans, miners & stealers (Windows + macOS) 🔸 Use hidden commands & indirect prompt injection Quick action: Never install random AI skills or models. Always verify the source. Read: thehackernews.com/2026/05/weekly…

@sunithvs_ @Cloudflare Yoi how can i get this?

Got $5000 @Cloudflare credits thanks @Cloudflare

@xoaanya You can burn 10M T/day can help you with that

Bought claude pro, how do I make 1 million per month??

@YashHustle_22 If you want a LAPtop then buy a mac

Give me one reason to buy MacBook in 2026.

@ritu_twts As an Ai engineer more the merries , else 24 or 32 be enough

As a developer, how much RAM is enough? 1. 4 GB 2. 8 GB 3. 16 GB 4. 24 GB 5. 32 GB 6. 64 GB+

@TeksEdge @basecampbernie When you combine this with agentic workfmoes there the promot processing tine increase sugnificantly due to the token contdxt length , unless we woukd have try with aider or some other agents

💡 Mind Blown - Shared Bandwidth Mini PCs run Medium Models! Mac Minis Are in Trouble? 💥$351 Minisforum UM790 Pro 🧩 AMD Radeon 780M iGPU (NO dedicated VRAM) 📦 48GB DDR5-5600 + 1TB NVMe (Added to ☠️) 🔥The secret? Unified memory + Vulkan + llama.cpp → iGPU borrows full system RAM (~75 GB/s bandwidth) → Allocates 21+ GB for model weights despite “4GB VRAM” label 🚀 Real results: • Gemma 4 26B MoE → 19.5 tok/s (196K context) • Gemma 4 E4B → 21.7 tok/s • Qwen3.5-35B-A3B → 20.8 tok/s • Nemotron Cascade 2 → 24.8 tok/s 🤖 Full local agentic workflows with @NousResearch Hermes + 40+ tools 🆓No API keys. Zero cost after purchase. 📍Fits in your hand. Runs forever.