GPT 5.5 Pro (2026-04-23)

released 2026-04-23

BenchAtlas Index

as of 2026-07-14

Not enough benchmark coverage for a composite score (needs ≥4 families across ≥3 categories). The per-benchmark evidence below stands on its own — sparse coverage is an evidence gap, not a low score.

Benchmark evidence

19 results

external indices

ECI	high effort	160.7 points	independent	Epoch AI Benchmarking Hub
ECI	xhigh effort	160.7 points	independent	Epoch AI Benchmarking Hub
ECI		160.7 points	independent	Epoch AI Benchmarking Hub

reasoning math

ARC-AGI-1older version split=public_eval · model_type=CoT	high effort	98.0 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	xhigh effort	98.0 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	high effort	96.5 %	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	xhigh effort	95.0 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	xhigh effort	90.4 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	high effort	90.1 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	high effort	84.6 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	xhigh effort	84.2 %	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	xhigh effort	11.1 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	xhigh effort	10.8 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=semi_private · model_type=CoT	high effort	10.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-2 split=public_eval · model_type=CoT	high effort	10.4 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	high effort	4.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=semi_private · model_type=CoT	xhigh effort	4.5 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	xhigh effort	3.3 usd_per_task	independent	ARC Prize Leaderboard
ARC-AGI-1older version split=public_eval · model_type=CoT	high effort	3.2 usd_per_task	independent	ARC Prize Leaderboard