Grok 4.20 (Reasoning): benchmark results
Grok 4.20 evaluated with reasoning enabled. Provider: xAI. Released 2026-03-10. Access: API.
Aggregate Rating 1677 ± 13, rank #429 of 2168 rated models, from 98 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Causal Sensitivity Score (CSS) | 47.3 | CSS (self-reported) | 100 |
| BenchTable - Utility | 87.4 | Weighted Score (%) | 99.6 |
| SpeechMap Compliance | 98.2 | % Requests Completed | 99.1 |
| BenchTable | 76.8 | Total Score (%) | 94.2 |
| CLBench - Rule System Application | 22.7 | Solving Rate (%) | 92.9 |
| Kagi LLM Benchmark | 75 | Accuracy (%) | 91.5 |
| AI Chess Leaderboard (Reasoning) | 1478 | Elo | 91.3 |
| Wolfram LLM Benchmarking Project | 66.3 | Correct Functionality (%) | 90.6 |
| AI for Education SEND | 83.94 | Accuracy (%) | 89.7 |
| CLBench | 22.2 | Solving Rate (%) | 88.6 |
| CLBench - Domain Knowledge Reasoning | 23.9 | Solving Rate (%) | 88.6 |
| BenchTable - STEM | 76.9 | Weighted Score (%) | 87.2 |
Interactive version: theaggregate.ai/model?slug=grok-4-20-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.