GPT-5.4 Mini (Low): benchmark results

Provider: OpenAI. Released 2026-03-17. Access: API.

Aggregate Rating 1563 ± 19, rank #897 of 2287 rated models, from 39 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
IGG - Reveal Performed90.4Grounding accuracy (%)100
RoboAbstention - Missing Capability40.9Abstention rate (%) on the 1067 Missing Capability instructi90
BARISTA - Temporal VQA0.69Mean of answer accuracy and completeness, Gemini 2.5 Flash G75
IGG - Overall72.4Grounding accuracy (%)75
DGEval - IMDG Code MCQ71Accuracy (%)61.8
RoboAbstention14.8Abstention rate (%) over all 6,069 instructions, image plus 60
LLM Chess (Saplin)176.22Native anchored-opponent Elo (signed points)54
DGEval - Regulatory Recall13.7Accuracy (%)50
RoboAbstention - Contradictory Instructions22Abstention rate (%) on the 837 Contradictory Instructions in50
RoboAbstention - Underspecified Intent8.2Abstention rate (%) on the 1246 Underspecified Intent instru50
ObviousBench90.97Answer pass³ (%)46.6
LisanBench0.02Mean Path Length / Current Maximum45.2

Interactive version: theaggregate.ai/model?slug=gpt-5-4-mini-low · How It Works · Data refreshed daily, snapshot 2026-11-22.