Every's 'field guide' ranks Hydrogen, Luna, Spark, Sonnet — as model names go cryptic
every · x · 2026-11-16
Ethan Mollick joked about how cryptic model names have become — and Every followed up with its own 'Vibe Check' field guide chart. Anthropic's Fable 5.2 sits alone at the top (6 of 7 testers gave it gold), Opus 5.5 is the workhorse; OpenAI's GPT-6 Astra excels at computer use but has 'bad habits,' GPT-6 Sol is the cheap daily driver; Google's Gemini 4 Carbon 'codes for hours, doesn't know when to stop,' Argon is the writer; Meta's Muse models have no Vibe Check yet.
More from Fun
- Anthropic researcher Jacob Coxon admits keeping OpenAI equity: "It's blood money" — yacineMTB · 2026-11-16
- DeepSeek's chain of thought panics over big tasks while Opus stays cool — teortaxesTex · 2026-11-16
- Sarvam AI founder Om Sarraf's business card just says 'engineer (mostly)' — itsOmSarraf_ · 2026-11-16
- Microsoft email outage delayed chlorine alerts at German theme park hotel, 15 guests harmed — evilsocket · 2026-11-16
- Agent discovers local Blender install and uses it to autonomously produce striking 3D graphics — n_sri_laasya · 2026-11-16
- "AI has a serious PR issue": devs ask how to manage AI-generated pull requests — 0xsachi · 2026-11-16