OpenAI GPT-5.5 Launch - Humanity's Last Exam (with tools): leaderboard

Metric: Score (%). Source: openai.com. 6 models tracked.

Top models

#ModelScore
1Claude Mythos Preview64.7
2Claude Mythos 564.5
3GPT-5.4 Pro (xHigh)58.7
4Claude Opus 4.857.9
5GPT-5.5 Pro57.2
6Claude Opus 4.754.7
7GPT-5.552.2
8GPT-5.452.1
9Gemini 3.1 Pro (Preview)51.4
10GPT-5.2 Pro50
11GPT-5.245.5

Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-5-5-launch-humanity-s-last-exam-with-tools · How It Works · Data refreshed daily, snapshot 2026-09-05.