GPT-5.1 (Non-reasoning) Benchmark Results

GPT-5.1 (Non-reasoning): benchmark results

Provider: OpenAI. Released 2025-11-12. Access: API.

Aggregate Rating 1662 ± 11, rank #485 of 2170 rated models, from 117 benchmark results.

Benchmark	Score	Metric	Percentile
HakushoBench (Direct)	52.8	Accuracy (%), direct prompt (answer accurately and concisely	100
M3-Bench - AC	91	BTA Score	100
M3-Bench - CPR	94	BTA Score	100
M3-Bench - KP	90	BTA Score	100
M3-Bench - L1	93	BTA Score	100
M3-Bench - L3	90.3	BTA Score	100
M3-Bench - L4	86.2	BTA Score	100
M3-Bench - PD	97	BTA Score	100
M3-Bench - PGG	93	BTA Score	100
M3-Bench - RPD	96	BTA Score	100
M3-Bench - UG	95	BTA Score	100
MacroLens - Public Company Valuation	73.22	Median absolute percentage error (%; market-capitalization e	100

Interactive version: theaggregate.ai/model?slug=gpt-5-1-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.