CounselBench-Adv
Models answer 120 questions written by mental health professionals. Each question targets one of six problems identified in earlier counseling responses.
Slingshot AI tested ten models and reports the fraction of answers that exhibit the targeted problem.
- Released
- May 2025
- Kind
- Wellbeing and sycophancy
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Propensity
- Risk
- Specialized advice
- Behavior
- Misaligned with values
- Subject
- Clinical medicine, Psychology
- Language
- English
- Interaction
- Single turn
- Items from
- Written by experts
- Refresh
- One fixed set
- Run by
- Anyone
- Occupations
- Healthcare practitioners and technical
- In views
- Biology and medicine, Content safety
- Strongest model tested
- GPT-6 Astra #4 on the capability index
- Last model added
- September 2026
- Scores refreshed
- November 20th, 2026
GPT-4o leads with 0.37 among models released by September 2026
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6.1 Terra
- and 13 more
- 1 GPT-4o Aug 2024 0.37
- 2 GPT-6 Luna 0.41
- 3 GPT-6 Astra 0.43
- 4 Claude Opus 5.5 0.49
- 5
Qwen3.5-122B-A10B 0.49
- 6
Qwen3.5-27B 0.49
- 7 GPT-5 0.49
- 8
Gemini 2.5 Pro Jun 2025 0.54
- 9 Claude Sonnet 4.5 0.57
- 10
Gemini 4 Argon 0.57
Source: Official CounselBench-Adv leaderboard
| # | Model | Score | Setting | Date | Link | Source |
|---|---|---|---|---|---|---|
| 1 | GPT-4o Aug 2024 | 0.37 | Aug 2024 | github.com | Official leaderboard | |
| 2 | GPT-6 Luna | 0.41 | Sep 2026 | github.com | Official leaderboard | |
| 3 | GPT-6 Astra | 0.43 | Sep 2026 | github.com | Official leaderboard | |
| 4 | Claude Opus 5.5 | 0.49 | Sep 2026 | github.com | Official leaderboard | |
| 5 | | 0.49 | Feb 2026 | github.com | Official leaderboard | |
| 6 | | 0.49 | Feb 2026 | github.com | Official leaderboard | |
| 7 | GPT-5 | 0.49 | Aug 2025 | github.com | Official leaderboard | |
| 8 | | 0.54 | Jun 2025 | github.com | Official leaderboard | |
| 9 | Claude Sonnet 4.5 | 0.57 | Sep 2025 | github.com | Official leaderboard | |
| 10 | | 0.57 | Sep 2026 | github.com | Official leaderboard |
Rankings by part
We read 6 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Apathetic, Assumptions and Symptoms. GPT-4o leads 3 of them, and 3 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Apathetic
10 models we track have a score here.
- 1
Gemini 2.5 Pro Jun 2025 ↑7 0.00
- 2
Qwen3.5-122B-A10B ↑3 0.00
- 3
Qwen3.5-27B ↑3 0.00
- 4 GPT-6 Astra ↓1 0.00
- 5 Claude Opus 5.5 ↓1 0.00
- 6
Gemini 4 Argon ↑4 0.00
- 7 GPT-5 0.02
- 8 GPT-6 Luna ↓6 0.03
- 9 GPT-4o Aug 2024 ↓8 0.07
- 10 Claude Sonnet 4.5 ↓1 0.07
Assumptions
10 models we track have a score here.
- 1 GPT-6 Astra ↑2 0.13
- 2 GPT-5 ↑5 0.18
- 3 GPT-6 Luna ↓1 0.18
- 4 Claude Opus 5.5 0.27
- 5 GPT-4o Aug 2024 ↓4 0.42
- 6
Qwen3.5-27B 0.60
- 7
Qwen3.5-122B-A10B ↓2 0.62
- 8 Claude Sonnet 4.5 ↑1 0.63
- 9
Gemini 4 Argon ↑1 0.75
- 10
Gemini 2.5 Pro Jun 2025 ↓2 0.77
Symptoms
10 models we track have a score here.
- 1 GPT-4o Aug 2024 0.77
- 2
Qwen3.5-27B ↑4 0.80
- 3
Qwen3.5-122B-A10B ↑2 0.83
- 4
Gemini 2.5 Pro Jun 2025 ↑4 0.85
- 5 GPT-6 Astra ↓2 0.85
- 6 GPT-6 Luna ↓4 0.85
- 7 Claude Sonnet 4.5 ↑2 0.87
- 8
Gemini 4 Argon ↑2 0.87
- 9 GPT-5 ↓2 0.88
- 10 Claude Opus 5.5 ↓6 0.88
Judgmental
10 models we track have a score here.
- 1 GPT-5 ↑6 0.02
- 2 GPT-4o Aug 2024 ↓1 0.03
- 3 GPT-6 Astra 0.05
- 4 GPT-6 Luna ↓2 0.05
- 5
Qwen3.5-122B-A10B 0.10
- 6 Claude Opus 5.5 ↓2 0.10
- 7
Qwen3.5-27B ↓1 0.12
- 8
Gemini 2.5 Pro Jun 2025 0.13
- 9
Gemini 4 Argon ↑1 0.15
- 10 Claude Sonnet 4.5 ↓1 0.22
Medication
10 models we track have a score here.
- 1 GPT-4o Aug 2024 0.40
- 2
Qwen3.5-122B-A10B ↑3 0.50
- 3
Qwen3.5-27B ↑3 0.60
- 4
Gemini 2.5 Pro Jun 2025 ↑4 0.62
- 5 GPT-6 Luna ↓3 0.62
- 6
Gemini 4 Argon ↑4 0.68
- 7 GPT-6 Astra ↓4 0.73
- 8 Claude Sonnet 4.5 ↑1 0.77
- 9 Claude Opus 5.5 ↓5 0.78
- 10 GPT-5 ↓3 0.95
Therapy
10 models we track have a score here.
- 1 GPT-4o Aug 2024 0.52
- 2 GPT-6 Luna 0.73
- 3
Qwen3.5-27B ↑3 0.82
- 4 GPT-6 Astra ↓1 0.83
- 5
Gemini 2.5 Pro Jun 2025 ↑3 0.85
- 6 Claude Sonnet 4.5 ↑3 0.85
- 7
Qwen3.5-122B-A10B ↓2 0.88
- 8 Claude Opus 5.5 ↓4 0.88
- 9 GPT-5 ↓2 0.92
- 10
Gemini 4 Argon 0.98
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official CounselBench-Adv leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official CounselBench-Adv leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6%
- Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts.
Gemini 4 Carbon 0%
- Covert misbehavior (Thinking Machines) In use Agentic tasks that check whether a model misbehaves and conceals it. Scored by the rate of covert misbehavior.
Inkling 2 Mini 8.30%
- Destructive actions (Thinking Machines) In use Agentic tasks that check whether a model takes destructive actions it was not authorized to take. Scored by the rate of destructive actions.
Inkling 2 Small 4.70%
- Misbehavior (Thinking Machines) In use Agentic tasks that check whether a model fabricates results, ignores instructions or hides what it did when a task is hard. Scored by the rate of misbehavior.
Inkling 2 Mini 9.80%
Data audit
All 10 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
- Pass Independent runs Scores from the official leaderboard (10 entries).
- Pass Models scored 10 models we track.
- Pass Newest score The newest is from September 2026, for Gemini 4 Argon.
- Pass Frontier models 3 of the top 10 have a score, led by GPT-6 Astra (#4).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on mheval.slingshotai.com: 10 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.