GPT-5.4 Mini: benchmark results

OpenAI's smaller, faster GPT-5.4 tier for cost-efficient reasoning workloads. Provider: OpenAI. Released 2026-03-17. Access: API.

Unified ELO 1658 ± 1, rank #113 of 1392 rated models, from 285 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AGC-Bench - hypobench0.48Dataset z-score100
CalBench149Excess (self-reported)100
GACL - Battleship85.94Normalized Score (0-100)100
TempGlitch52.4Acc. (1 FPS) (self-reported)100
AGC-Bench - pun_eval1.4Dataset z-score97.6
AGC-Bench - humor_transfer0.97Dataset z-score96.3
LLM-as-a-Reviewer99.7Low (NeurIPS 2022) (self-reported)95.5
AGC-Bench - pollux_creativity0.92Dataset z-score95.1
SLR-Bench - Hard28Accuracy (%)94.5
Roboflow Vision Evals - Visual Understanding77.61Pass Rate (%)93.4
Chartographer94ChartQA OA (self-reported)92.9
AGC-Bench - pron_vs_prompt1.06Dataset z-score92.6

Interactive version: theaggregate.ai/model?slug=gpt-5-4-mini · How It Works · Data refreshed daily, snapshot 2026-09-05.