Grok 4.7 tipped as 2.1T model, but benchmarks show no token-efficiency gain over 4.6
teortaxesTex · x · 2026-09-22
Quoting theo's report that Grok 4.7 will be a 2.1T-parameter model beating 4.6 across the board (slower to serve but more token-efficient), teortaxesTex argues xAI follows Elon's rapid-rebuild playbook — shipping base models fast but leaving them under-debugged, with 4.6 a step up from 4.5 while 4.7 starts from scratch. He notes no benchmark yet shows 4.7 more token-efficient than 4.6.
More from Models
- Dev visualizes LFM2.5-VL driving decisions from first-token probabilities — helloiamleonie · 2026-09-22
- Open-weight AI is basically Chinese labs: token share data shows near-total dominance — gaganghotra_ · 2026-09-22
- OpenAI model trained for just 24 days reportedly solved 100+ longstanding open math problems — FinanceYF5 · 2026-09-22
- Vacation Project: Open-Sourcing a 0.2B Arabic Model Trained for $100.63 in 30 Hours — assemsabryy · 2026-09-22
- Muse model demoed running on Meta Glasses, tipped as big Connect focus — Scobleizer · 2026-09-22
- Peking University releases OmniEdu, open education foundation models in 4B/9B/27B sizes — PekingUniversity · 2026-09-22