OpenAI GPT-5.4-Pro vs xAI Grok 3 — side-by-side specs.
OpenAI GPT-5.4-Pro | xAI Grok 3 | |
|---|---|---|
| Overview | ||
| Company | OpenAI | xAI |
| Release date | Mar 5 2026 | Feb 17 2025 |
| Model type | — | — |
| Open source | No | No |
| Specifications | ||
Parameters | — | — |
Context window | — | — |
| Benchmarks | ||
Web browsing BrowseCompCan the AI browse the web and track down hard-to-find answers? Higher is better. | 89.3% | — |
Advanced math FrontierMath · Tier 1–3Very hard, research-level math problems. Tiers 1–3 are the (still extremely difficult) lower tiers. Higher is better. | 50% | — |
Advanced math FrontierMath · Tier 4Very hard, research-level math problems. Tier 4 is the hardest — close to what professional research mathematicians tackle. Higher is better. | 38% | — |
Science GPQA DiamondGraduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 94.4% | — |
Knowledge work GDPval (win/tie rate)How often the AI's work matches or beats a human expert's on real knowledge-work tasks. Higher is better. | 82% | — |
| Timeline | ||
| Release gap | Grok 3 shipped 381 days before GPT-5.4-Pro | |
GPT-5.4-Pro and Grok 3 don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Grok 3 shipped 381 days before GPT-5.4-Pro, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
Direct benchmark comparisons are unavailable — GPT-5.4-Pro and Grok 3 don't publish scores on any of the same benchmarks.