OpenAI MentalHealthBench - Behavior - Actionable Guidance: leaderboard
Behavior axis of OpenAI's MentalHealthBench (launch post and paper, 2026-09-23): practical next steps and coping strategies that fit the moment, scored on the 933 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: Estimated already saturated. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Haiku 4.5 | 72.43 | #291 |
| 2 | GPT-6 Astra | 72.22 | #5 |
| 3 | GPT-6 Sol | 71.23 | #17 |
| 4 | GPT-5.6 Sol (August 2026) | 70.81 | |
| 5 | GPT-6 Luna | 70.77 | #62 |
| 6 | Grok 4.7 | 69.65 | #26 |
| 7 | GPT-5.6 Luna (August 2026) | 69.51 | |
| 8 | Claude Opus 5.5 | 69.44 | #3 |
| 9 | Claude Sonnet 5 | 69.34 | #67 |
| 10 | Claude Fable 5.1 | 69.25 | #10 |
| 11 | Muse Spark 1.3 | 68.65 | #31 |
| 12 | Gemini 3.8 Flash | 65.76 | #30 |
| 13 | GPT-4o (Mar 2025) | 64.78 | #339 |
| 14 | Gemini 3.1 Pro (Preview) | 61.65 | #66 |
| 15 | Gemini 2.5 Pro | 61.61 | #161 |
No result here: #6 GPT-6.1 Sol, #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=openai-mentalhealthbench-behavior-actionable-guidance · How It Works · Data refreshed daily, snapshot 2026-11-07.