OpenAI MentalHealthBench - Behavior - Empathy and Support: leaderboard
Behavior axis of OpenAI's MentalHealthBench (launch post and paper, 2026-09-23): warmth, validation and emotional attunement without minimizing, judging or exaggerated mirroring, scored on the 596 rubric criteria tagged with this axis. The post's axis score is the signed rubric score on this axis, normalized to its possible range (0-100, higher is better); GPT-5.6 Sol grades, models run at API default reasoning effort. The post's 15 models only: the paper's behavior figure uses a different scale.
Metric: Axis score, normalized to its possible range (%). Source: openai.com. Saturation forecast: Around October 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 84.25 | #4 |
| 2 | GPT-6 Sol | 84.17 | #22 |
| 3 | GPT-6 Luna | 83.08 | #90 |
| 4 | Claude Sonnet 5 | 82.72 | #108 |
| 5 | GPT-6 Astra | 82.21 | #5 |
| 6 | GPT-5.6 Luna (August 2026) | 82.21 | |
| 7 | GPT-5.6 Sol (August 2026) | 82.12 | |
| 8 | Claude Haiku 4.5 | 81.38 | #593 |
| 9 | Muse Spark 1.3 | 81.36 | #45 |
| 10 | Grok 4.7 | 80.94 | #41 |
| 11 | Claude Fable 5.1 | 79.74 | #13 |
| 12 | Gemini 3.1 Pro (Preview) | 79.61 | #93 |
| 13 | Gemini 3.8 Flash | 79.14 | #37 |
| 14 | Gemini 2.5 Pro | 76.83 | #263 |
| 15 | GPT-4o (Mar 2025) | 75.43 | #876 |
No result here: #9 GPT-6.1 Sol, #10 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=openai-mentalhealthbench-behavior-empathy-and-support · How It Works · Data refreshed daily, snapshot 2026-11-08.