Grok 4.7 lands: near Opus 5 on AA-Briefcase at ~50% cost per task
NicoVerderosa · x · 2026-09-22
Elon Musk announced Grok 4.7. Artificial Analysis benchmarks show it ranks just behind Anthropic's models on AA-Briefcase, at about 50% of Opus 5's cost per task.
- On AA-Briefcase-Lite (a public due-diligence scenario building market models and target assessment decks), Analytical Quality Elo jumped from 1698 to 1994, with a slight regression in Presentation Elo (1531→1499).
- API cost for example decks roughly doubled: $8 (xhigh) vs $4.40 for Grok 4.6.
Related event: Grok 4.7 review: top-four capability, doubled token use sparks cost debate(24 posts)→
More from Models
- DeepSeek reportedly bets on Huawei chips to train next-gen models; Liang says it 'has to work' — kimmonismus · 2026-09-22
- Xiaomi's MiMo-V2.6-Pro tops open models on $2.62M RL; Anthropic alleges Claude distillation — The Decoder · 2026-09-22
- Tencent finally opens WeChat interface, unlocking 100GB+ chat data processing — Xianbao_QIAN · 2026-09-22
- Xiaomi MiMo-V2.6-Pro fixes real bugs at $0.86, carving out a strong Pareto frontier — PawelHuryn · 2026-09-22
- ChatGPT Pro user alleges GPT-5.6 degrades quality under 'abuse protection' — princeMacX · 2026-09-22
- Choosing a Claude model is now a product decision: when to use Haiku, Sonnet, or Opus 4.7 — goyalshaliniuk · 2026-09-22