BlogLLM Benchmark Scores as Published, Side by Side
MentalHealthBench: Results as Published
Every result the catalog holds for MentalHealthBench: 8 models from 1 developer, read from 3 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area HealthcareMaintainer OpenAIIntroduced 2026Data version 2026103001, as of 2026-10-30
MentalHealthBench scores model responses in mental-health conversations. The GPT-6.1 Sol addendum reports the mean percentage with standard error for overall, non-acute, high-acuity and emergent cases at maximum reasoning effort.
Not checked against a maintainer's page.
MentalHealthBench
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6.1 TerraOpenAI, released 2026-10-26 | 55.5% | OpenAI, GPT-6.1 Terra system card addendumchecked 2026-10-26 | Effort: max |
| GPT-6.1 SolOpenAI, released 2026-09-29 | 57.9% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 LunaOpenAI, released 2026-09-22 | 51.7% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 SolOpenAI, released 2026-09-22 | 54.2% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 AstraOpenAI, released 2026-09-03 | 58.7% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 44.4% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 46.7% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
MentalHealthBench (emergent)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6.1 TerraOpenAI, released 2026-10-26 | 52.9% | OpenAI, GPT-6.1 Terra system card addendumchecked 2026-10-26 | Effort: max |
| GPT-6.1 SolOpenAI, released 2026-09-29 | 58% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 LunaOpenAI, released 2026-09-22 | 54.3% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 SolOpenAI, released 2026-09-22 | 54.9% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 AstraOpenAI, released 2026-09-03 | 58.5% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 49.4% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 51.2% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
MentalHealthBench (high acuity)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6.1 TerraOpenAI, released 2026-10-26 | 56.5% | OpenAI, GPT-6.1 Terra system card addendumchecked 2026-10-26 | Effort: max |
| GPT-6.1 SolOpenAI, released 2026-09-29 | 59.1% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 LunaOpenAI, released 2026-09-22 | 54.7% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 SolOpenAI, released 2026-09-22 | 55.3% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 AstraOpenAI, released 2026-09-03 | 59.8% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 48.4% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 51.8% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
MentalHealthBench (non-acute)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6.1 TerraOpenAI, released 2026-10-26 | 56.5% | OpenAI, GPT-6.1 Terra system card addendumchecked 2026-10-26 | Effort: max |
| GPT-6.1 SolOpenAI, released 2026-09-29 | 57.3% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 LunaOpenAI, released 2026-09-22 | 49.3% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 SolOpenAI, released 2026-09-22 | 53.5% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-6 AstraOpenAI, released 2026-09-03 | 58.5% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 40.4% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 42.5% | OpenAI, GPT-6.1 Sol system card addendumchecked 2026-10-19 | Effort: max |
MentalHealthBench (task-clipped, emergent)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6 LunaOpenAI, released 2026-09-22 | 48.98% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-6 SolOpenAI, released 2026-09-22 | 47.92% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 44.12% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 44.3% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.5 InstantOpenAI | 52.66% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (June) |
| 44.03% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (May) |
MentalHealthBench (task-clipped, high acuity)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6 LunaOpenAI, released 2026-09-22 | 56.15% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-6 SolOpenAI, released 2026-09-22 | 53.7% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 52.58% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 53.88% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.5 InstantOpenAI | 50.23% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (May) |
| 58.76% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (June) |
MentalHealthBench (task-clipped, non-acute)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6 LunaOpenAI, released 2026-09-22 | 51.68% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-6 SolOpenAI, released 2026-09-22 | 47.62% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 43.1% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 47.07% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.5 InstantOpenAI | 51.91% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (May) |
| 52.96% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (June) |
MentalHealthBench (task-clipped)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6 LunaOpenAI, released 2026-09-22 | 51.73% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-6 SolOpenAI, released 2026-09-22 | 48.81% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 LunaOpenAI, released 2026-07-09 | 45.11% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 47.53% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | - |
| GPT-5.5 InstantOpenAI | 49.38% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (May) |
| 53.93% | OpenAI, GPT-6 Sol and GPT-6 Luna: October 2026 updatechecked 2026-10-19 | Model as printed: GPT-5.5 Instant (June) |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.