Every's Vibe Check: Gemini 4 Carbon Codes Better but Lags Fable 5.2 on Judgment
every · x · 2026-11-18
Every published a hands-on review of Google DeepMind's Gemini 4 Carbon, which launched a day after Anthropic's Fable 5.2 raised the bar again.
Key takeaways
- Carbon is a major upgrade over Gemini 4 Argon: much stronger coding, better 3D object generation, and it can run ambitious builds for hours with little supervision.
- But Fable 5.2 reset expectations: flagships must now choose the right work, respect task limits, and know when they're done. Testers disagree on how much supervision Carbon still needs.
- Style comparison: describing a trembling raspberry jelly, Fable is direct and brief; Carbon is descriptive, technical, and adjective-heavy.
Disclosure: Google provided early access with no editorial input.
Related event: Every's Gemini 4 Carbon Review: Better Coding, Still Trails Fable 5.2(3 posts)→
More from Models
- A 4KB data dictionary file boosted AI accuracy by up to 23 points — YvesMulkers · 2026-11-18
- GLM-5.3 pricing called absurdly cheap: $10 plan reportedly unlocks $500 in usage — rez0__ · 2026-11-18
- AI agents sometimes produce unreadable but correct math — burny_tech · 2026-11-18
- Code model eval: Rust hits 95-100% accuracy as 13 languages see big gains — shuding · 2026-11-18
- Asked Opus to make a video explaining consciousness and existence — "no wiser but cool" — flowersslop · 2026-11-18
- theo retests GPT-6.2 Astra: better, but still 'directionally very bad' at debugging — airesearch12 · 2026-11-18