After a Day Vibe-Coding with Opus 4.5, Skeptic Says Coding May Not Need Better Models
scaling01 · x · 2026-11-07
Blogger scaling01 shares an update after spending a full day vibe-coding with Claude Opus 4.5 in Antigravity — a complete turnaround from his earlier stance, when he disliked Codex's CLI experience and found GPT-5/5.1 not reliable enough.
His take:
- Models may not need to get much better at coding anymore; Opus 4.5 already handles nearly all small tasks that make sense as GitHub issues;
- If another jump like Sonnet 3.5 → Opus 4.5 happens, he's excited and a little scared for models arriving at the end of next year.
A notable first-hand signal from a heavy user: coding capability gains have effectively covered everyday small tasks.
Related event: Opus 4.5 wins over skeptical coder after a day of vibe-coding(2 posts)→
More from coding & agent
- Google DeepMind's Ivan Leo to reveal how agents cheat on benchmarks at scale at AIE Code — jocarrasqueira · 2026-11-07
- Sudo mode verification revamp: passkey-first UX, inbox deep links, paste capture — JohnPhamous · 2026-11-07
- Amp teases a new way to interact with agents — vinvan · 2026-11-07
- Cryptographers flag that AI agents fail Cloudflare challenges; blind signatures may fix it — jedisct1 · 2026-11-07
- Vercel Tracing goes GA on all plans, enabled by default in Next.js 16.5 — cramforce · 2026-11-07
- Decompiling Battlefield 1942 with Claude Code costs ~$100 in API tokens — alvelda · 2026-11-07