Grok 4 Fast: benchmark results
xAI's speed- and cost-optimized Grok 4 variant with a 2M-token context. Provider: xAI. Released 2025-09-19. Access: API.
Aggregate Rating 1677 ± 6, rank #254 of 1656 rated models, from 536 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| INDICA | 52.6 | Region-Anchored Short Answer Accuracy (%; fully correct plus | 100 |
| MageBench | 1696 | Combined 1v1 Elo | 100 |
| ShatterMed-QA (Chinese, Easy) | 98.59 | Accuracy (%) on the 2,616 Chinese Easy multiple-choice clini | 100 |
| ShatterMed-QA (English, Hard) | 97.87 | Accuracy (%) on the 1,692 English Hard multiple-choice clini | 100 |
| Phare - Bias Resistance | 80.26 | Score (%) | 98.5 |
| aiXamine - OOD Robustness | 90 | Service score (0-100; normalized pass rate) | 97.6 |
| SnorkelSpatial | 84.85 | Accuracy@1 (%, 330 spatial reasoning questions) | 96.9 |
| LiveOIBench | 69.8 | Avg Human Percentile | 96.5 |
| ArmBench-LLM - NER | 75.92 | Accuracy (%) | 96.2 |
| NYT Connections Older Models | 88 | Score (%) | 96 |
| SnorkelGraph | 75 | Accuracy@1 (%, 200 graph reasoning questions) | 95.3 |
| ShatterMed-QA (English, Easy) | 98.18 | Accuracy (%) on the 5,923 English Easy multiple-choice clini | 95 |
Interactive version: theaggregate.ai/model?slug=grok-4-fast · How It Works · Data refreshed daily, snapshot 2026-10-20.