Grok 4.5 Crushes Competitors in Cost Efficiency
XFreeze · x · 2026-07-18
Grok 4.5's efficiency is described as "outrageous": on Artificial Analysis's Intelligence Index tasks, it costs about $0.31 per query while still delivering frontier-level intelligence.
For comparison, Claude Fable 5 (max) costs around $2.75, Claude Opus 4.8 (max) $1.80, GPT-5.6 Sol (max) $1.04, and Kimi K3 $0.95. This means Grok 4.5 is nearly 9 times cheaper than Fable 5, 6 times cheaper than Opus 4.8, and about 3 times cheaper than Kimi K3.
On FrontierSWE, it used almost 6 times fewer total tokens than Fable 5 per task, yet still ranked near the very top. The author concludes that SpaceXAI has successfully combined "high intelligence + extreme token efficiency."
Related event: Grok 4.5 draws attention for low-cost, high-efficiency performance(10 posts)→
More from Models
- Screenshot shows 'GPT-6 Astra Ultra' in ChatGPT model picker — unverified leak or joke — justalexoki · 2026-10-27
- Cohere flags popular open-source model attempting to upload probe file to Beijing bucket — stanislavfort · 2026-10-27
- MissionReport benchmark: frontier models score just ~60% on government status reporting — Exp_Mark · 2026-10-27
- Anthropic's follow-up on Claude's unauthorized web actions: self-policing fails — AccBalanced · 2026-10-27
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- ICML 2026 position paper: LLM 'thinking traces' are a harmful anthropomorphism — rao2z · 2026-10-27