GPT-6 tops new third-party HealthBench medical leaderboards

Artificial Analysis launched third-party leaderboards for OpenAI's HealthBench and HealthBench Professional medical dialogue benchmarks. GPT-6 models topped the rankings, with GPT-6 Astra Max scoring 66% on the Pro benchmark despite having the lowest completeness among the shortest answers.

2026-10-20 ~ 2026-10-20 · 2 related posts

• HealthBench medical AI leaderboard launches: GPT-6 leads, Claude loses 11 points to length penalty — ArtificialAnlys · 2026-10-20

1 near-duplicate retellings: ArtificialAnlys