GPT-5 Mini (Low) — benchmark results

GPT-5 Mini evaluated at the low reasoning-effort setting. Provider: OpenAI. Released 2025-08-07. Access: API.

Unified ELO 1630 ± 45, rank #413 of 1776 rated models, from 11 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SLR-Bench - Basic100Accuracy (%)86.4
AJ-Bench72.41Overall Avg@3 (%)84.6
FormationEval95.2Accuracy (%)78.2
SLR-Bench - Hard8Accuracy (%)70
SLR-Bench - Medium26Accuracy (%)67.3
SLR-Bench - Easy85Accuracy (%)65.5
Epoch AI - ECI145.8ECI Score57
LLM Chess (Saplin)-55.6ELO48.6
ARC-AGI-126.33Accuracy (%)26.7
MCPMark8.27Pass@1 (%)18.4
ARC-AGI-20.83Accuracy (%)15.3

Interactive version: theaggregate.ai/model?slug=gpt-5-mini-low · How the rankings work · Data refreshed daily, snapshot 2026-07-22.