Benchgen

HealthBench — Results

RankModelScore
1gpt-6-astra63.4
2kimi-k2-thinking-09050.58
3gpt-oss-120b-high0.576
4gpt-5-6-sol0.57
5gpt-5-6-terra0.57
6gpt-5-6-luna0.558
7gpt-5-3-codex0.541
8gpt-5-50.514

HealthBench — Phases

1 phaseActive

Phase 1·

HealthBench

Aug 3, 2026
Current