GPT-5.6 Terra on HealthBench Professional

rank 5 of 9 · updated August 16, 2026

GPT-5.6 Terra scores 0.577 on HealthBench Professional, rank 5 of 9 evaluated models. The balanced middle tier of the GPT-5.6 family, between Sol and Luna. HealthBench Professional scores models on 525 tasks drawn from real clinician conversations, graded criterion by criterion against physician-written rubrics on a 0 to 1 scale.

Result and API facts

Position in the field

The gap to the leader, Claude Fable 5 at 0.660, is 0.083. Directly above sits Claude Sonnet 5 at 0.578. Directly below sits Claude Opus 4.8 at 0.558. Scores on this page come from the same evaluation run, so differences between models are differences on identical tasks, not across configurations.

Head to head

Pairings with a dedicated comparison page are linked; every other difference is in the score-difference matrix.

• GPT-5.6 Terra vs Claude Fable 5
  0.577 vs 0.660 · Claude Fable 5 by 0.083
• GPT-5.6 Terra vs GPT-5.6 Sol
  0.577 vs 0.605 · GPT-5.6 Sol by 0.028
• GPT-5.6 Terra vs Claude Opus 5
  0.577 vs 0.598 · Claude Opus 5 by 0.021
• GPT-5.6 Terra vs Claude Sonnet 5
  0.577 vs 0.578 · Claude Sonnet 5 by 0.001
• GPT-5.6 Terra vs Claude Opus 4.8
  0.577 vs 0.558 · GPT-5.6 Terra by 0.019
• GPT-5.6 Terra vs GPT-5.6 Luna
  0.577 vs 0.557 · GPT-5.6 Terra by 0.020
• GPT-5.6 Terra vs GPT-5.5 Instant
  0.577 vs 0.384 · GPT-5.6 Terra by 0.193
• GPT-5.6 Terra vs MAI-Thinking-1
  0.577 vs 0.350 · GPT-5.6 Terra by 0.227

How tasks are selected and graded is on the methodology page. The full ranking is on the leaderboard.