@marc_illy: Anthropic just dropped Claude Opus 4.8 and nobody’s talking about it 👀🤖 Same price. Smarter. Faster. More honest. 🔥 Here’s what actually changed 👇 ✅ 69.2% on SWE-bench (coding benchmark) — up from 64.3% ✅ 4x less likely to let code bugs slide without flagging them ✅ Massive math jump — 96.7% on USAMO 2026 vs 69.3% on 4.7 ✅ Works independently for longer on complex tasks ✅ Flags its own uncertainty instead of hallucinating Still $5/million input tokens. Same price as 4.7. 💀 And Anthropic just teased that Mythos — their most powerful model ever — drops “in the coming weeks” 👁️ We’re not in the early days of AI anymore. This is a new model dropping every 6 weeks. If you’re still sleeping on this… your competition isn’t 🫡 #claude #opus48 #anthropic #aitools #artificialintelligence