MAI-Thinking-1 on HealthBench Professional

rank 9 of 9 · updated August 16, 2026

MAI-Thinking-1 scores 0.350 on HealthBench Professional, rank 9 of 9 evaluated models. Microsoft AI's first in-house frontier reasoning model, a sparse mixture-of-experts design in public preview on Microsoft Foundry. 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 0.310. Directly above sits GPT-5.5 Instant at 0.384. 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.

• MAI-Thinking-1 vs Claude Fable 5
  0.350 vs 0.660 · Claude Fable 5 by 0.310
• MAI-Thinking-1 vs GPT-5.6 Sol
  0.350 vs 0.605 · GPT-5.6 Sol by 0.255
• MAI-Thinking-1 vs Claude Opus 5
  0.350 vs 0.598 · Claude Opus 5 by 0.248
• MAI-Thinking-1 vs Claude Sonnet 5
  0.350 vs 0.578 · Claude Sonnet 5 by 0.228
• MAI-Thinking-1 vs GPT-5.6 Terra
  0.350 vs 0.577 · GPT-5.6 Terra by 0.227
• MAI-Thinking-1 vs Claude Opus 4.8
  0.350 vs 0.558 · Claude Opus 4.8 by 0.208
• MAI-Thinking-1 vs GPT-5.6 Luna
  0.350 vs 0.557 · GPT-5.6 Luna by 0.207
• MAI-Thinking-1 vs GPT-5.5 Instant
  0.350 vs 0.384 · GPT-5.5 Instant by 0.034

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