7 testers put Gemini 4 Carbon through a Reach Test: 2 would always use it, 1 won't
every · x · 2026-11-11
Every had 7 team members test Gemini 4 Carbon for weeks and ran a "Reach Test" to see who would actually keep using it.
Results: 2 would definitely reach for it, 4 would use it sometimes, 1 wouldn't.
- Kieran Klaassen (coding): values its tests and docs on large coding projects; ranks Google second among coding labs, behind Anthropic
- Randy Counsman (video): daily driver for weeks; caught audio/video problems other models missed
- Jack Cheng (editing): good for icons and 3D objects, not pages
- Katie Parrott (writing): a better builder than Argon but a worse writer
- Others: creative UI ideas but weaker at copying real objects; writes about like Argon but costs more
Takeaway: a model can improve substantially without becoming everyone's first choice.
More from Models
- Alleged Gemini 4 Carbon builds 92-part 3D Paris physics lab from one image (unverified) — theREALmarvin · 2026-11-11
- GBENCH: OpenAI's GPT-6 Astra tops unsaturated coding benchmark with 83.3, sweeping top 3 — teortaxesTex · 2026-11-11
- Google launches Gemini 4 Carbon, its strongest model for long-horizon multi-step reasoning — laparisa · 2026-11-11
- Artificial Analysis launches Intelligence Index v5 at San Francisco event — ArtificialAnlys · 2026-11-11
- Fable 5.2 video generation via code appears to lack nudity filters in JS — mickeyxfriedman · 2026-11-11
- Lenny: Designers name Claude most often, Anthropic winning hearts and minds — lennysan · 2026-11-11