Enter opens the highlighted result. Esc closes.

MentalHealthBench (high acuity) leaderboard

As of November 22, 2026, among the results listed here, OpenAI GPT-6 Astra has the highest reported MentalHealthBench (high acuity) score, 59.8 (high acuity), published by OpenAI, ahead of OpenAI GPT-6.1 Sol at 59.1. 10 models have a MentalHealthBench (high acuity) result in our records.

Scores as the model makers published them, not re-run by us; each row links to its announcement.

Best reported score

59.8

OpenAI GPT-6 Astra

Reported by OpenAI, September 4, 2026 · high acuity

Runner-up

59.1

OpenAI GPT-6.1 Sol

Reported by OpenAI, September 4, 2026 · high acuity

10 models ranked · 1 maker

Reported MentalHealthBench (high acuity) scores

MentalHealthBench (high acuity) scores reported by model makers
Rank Model Score Setting Reported by Date
1 GPT-6 Astra 59.8 high acuity OpenAI September 4, 2026 (model release)
2 GPT-6.1 Sol 59.1 high acuity OpenAI September 4, 2026 (model release)
3 GPT-6.1 Terra 56.5 high acuity OpenAI October 26, 2026 (model release)
4 GPT-6.2 Astra 56.4 high-acuity OpenAI November 17, 2026 (model release)
5 GPT-6 Sol 55.3 high acuity OpenAI September 4, 2026 (model release)
6 GPT-6 Luna 54.7 high acuity OpenAI September 4, 2026 (model release)
7 GPT-6.2 Sol 52.9 high-acuity OpenAI November 17, 2026 (model release)
8 GPT-5.6 Sol 51.8 high acuity OpenAI September 4, 2026 (model release)
9 GPT-5.6 Terra 51.3 high acuity OpenAI October 26, 2026 (model release)
10 GPT-5.6 Luna 48.4 high acuity OpenAI September 4, 2026 (model release)

Other reported runs

MentalHealthBench (high acuity) scores reported by model makers, other reported runs
Model Score Setting Reported by Date
GPT-6 Astra 59.8 high-acuity OpenAI November 17, 2026 (model release)
GPT-6 Sol 55.3 high-acuity OpenAI November 17, 2026 (model release)
GPT-6.1 Sol 59.1 high-acuity OpenAI November 17, 2026 (model release)
GPT-6.1 Terra 56.5 high-acuity OpenAI November 17, 2026 (model release)

How much to trust these numbers

Source
Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
Settings
Settings differ between makers. Check the Setting column before comparing two scores.
Spread
59.8 to 48.4 across 10 models.
Freshness
Newest result published November 17, 2026; oldest September 4, 2026.

What MentalHealthBench (high acuity) measures

MentalHealthBench (high acuity) appears in the release announcements of OpenAI. We have not found a public write-up of the benchmark itself yet.