Subscribe
THE DOSSIER · ALL ENTITIES

MentalHealthBench

The permanent record for MentalHealthBench. Every claim we have put on the ledger, checked against independent sources, with the verdict and the date. On the record so far, MentalHealthBench is on the record, but the sample is too small to call a pattern. This page compounds: each new verdict updates the rate.

40BS RATE / 100
of the 1 claim we actually checked1
0 VERIFIED 1 TRUE, BUT 0 BS
MOST RECENT ON THE RECORDOpenAI built a mental health test, had its own model grade it, and its own model came top. Answers written by licensed clinicians scored 38.5%. GPT-6 Astra scored 57.3%.TRUE, BUT 25 Sept 2026 · 13 SOURCES
THE FULL RECORD · 1 CLAIMS, NEWEST FIRST
TRUE, BUTOpenAI built a mental health test, had its own model grade it, and its own model came top. Answers written by licensed clinicians scored 38.5%. GPT-6 Astra scored 57.3%.CAPABILITY · 25 Sept 2026 · 13 SOURCES

SEARCH EVERY MENTALHEALTHBENCH CLAIM IN THE LEDGER →
← ALL ENTITY DOSSIERS

---