What happened
Artificial Analysis is an independent group that tests AI models. It scored Grok 4.5 at 54 on its Intelligence Index. That put Grok 4.5 fourth overall. On a second test, GDPval-AA v2, it also ranked fourth. Its result landed just behind Anthropic's Opus 4.8 and ahead of a model called GLM-5.2.
The context
Fourth place is a solid result for xAI. It trails the top models from Anthropic, OpenAI, and Google, but stays close.
What this means for you
👤EverydayIndependent scores help you judge which AI is strong, apart from company claims.
💼At workA top-five model is a safe pick for demanding tasks like coding and analysis.
🏢BusinessOutside benchmarks give buyers a neutral way to compare before spending.
🌍The worldPublic leaderboards keep every lab honest about how good their models really are.
Ledger entry: Capability · Materiality 3 · T2 Measured → bounded ▲ for xAI
Sources
- Artificial Analysis, https://artificialanalysis.ai/models/grok-4-5 T2 Measured