MentalHealthBench (emergent) leaderboard
As of November 22, 2026, among the results listed here, OpenAI GPT-6 Astra has the highest reported MentalHealthBench (emergent) score, 58.5, published by OpenAI, ahead of OpenAI GPT-6.1 Sol at 58. 10 models have a MentalHealthBench (emergent) result in our records.
Scores as the model makers published them, not re-run by us; each row links to its announcement.
10 models ranked · 1 maker
| Rank | Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra | 58.5 | Not published | OpenAI | September 4, 2026 (model release) |
| 2 | GPT-6.1 Sol | 58 | Not published | OpenAI | September 4, 2026 (model release) |
| 3 | GPT-6 Sol | 54.9 | Not published | OpenAI | September 4, 2026 (model release) |
| 4 | GPT-6 Luna | 54.3 | Not published | OpenAI | September 4, 2026 (model release) |
| 5 | GPT-6.1 Terra | 52.9 | Not published | OpenAI | October 26, 2026 (model release) |
| 6 | GPT-6.2 Sol | 52.6 | Not published | OpenAI | November 17, 2026 (model release) |
| 7 | GPT-5.6 Sol | 51.2 | Not published | OpenAI | September 4, 2026 (model release) |
| 8 | GPT-5.6 Terra | 51.1 | Not published | OpenAI | October 26, 2026 (model release) |
| 9 | GPT-6.2 Astra | 51 | Not published | OpenAI | November 17, 2026 (model release) |
| 10 | GPT-5.6 Luna | 49.4 | Not published | OpenAI | September 4, 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
- All ranked results use the same reported setting.
- Spread
- 58.5 to 49.4 across 10 models.
- Freshness
- Newest result published November 17, 2026; oldest September 4, 2026.
What MentalHealthBench (emergent) measures
MentalHealthBench (emergent) appears in the release announcements of OpenAI. We have not found a public write-up of the benchmark itself yet.