GPT-5 Nano (Low) — benchmark results
GPT-5 Nano evaluated at the low reasoning-effort setting. Provider: OpenAI. Released 2025-08-07. Access: API.
Unified ELO 1440 ± 60, rank #1040 of 1776 rated models, from 13 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SLR-Bench - Basic | 100 | Accuracy (%) | 86.4 |
| FormationEval | 93.3 | Accuracy (%) | 57 |
| Epoch AI - ECI | 140.86 | ECI Score | 38.8 |
| LLM Chess (Saplin) | -117 | ELO | 38.6 |
| SLR-Bench - Easy | 48 | Accuracy (%) | 18.2 |
| WeirdML | 25.89 | Average Score | 16.8 |
| SLR-Bench - Hard | 1 | Accuracy (%) | 15.5 |
| MathArena - AIME 2025 | 47.22 | Accuracy (%) | 12.9 |
| SLR-Bench - Medium | 2 | Accuracy (%) | 11.8 |
| MCPMark | 4.33 | Pass@1 (%) | 5.3 |
| ARC-AGI-2 | 0 | Accuracy (%) | 4.9 |
| MageBench Season 1 | 1499 | Rating (self-reported) | 2.9 |
Interactive version: theaggregate.ai/model?slug=gpt-5-nano-low · How the rankings work · Data refreshed daily, snapshot 2026-07-22.