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.

Unified ELO 1627 ± 1, rank #189 of 1392 rated models, from 73 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SnakeBench33.2TrueSkill Rating95.1
GSMA Open-Telco - TeleLogs75Score (%)93.2
GSMA Open-Telco - TeleQnA83.6Score (%)92.6
YapBench70.2YapIndex (lower is better)90.3
Diplomacy: Betrayal Tendency0Betrayal Rate (%)87.5
LOL Arena47.1Win Rate (%)87.5
GSMA Open-Telco - srsRAN-Bench84Score (%)86.9
GSMA Open-Telco LLM Leaderboard68.42Average Score (%)85.2
Fiction.LiveBench75Accuracy (%)82.1
GSMA Open-Telco - ORAN-Bench83.33Score (%)81.8
FormationEval96Accuracy (%)81.7
Chess Bench LLM1220Lichess Rating81.1

Interactive version: theaggregate.ai/model?slug=grok-4-fast · How It Works · Data refreshed daily, snapshot 2026-09-05.