Claude Opus 4.8 (Opus 4.8 grader) on HealthBench Professional

rank 15 of 31 · updated September 30, 2026

Claude Opus 4.8 (Opus 4.8 grader) scores 0.574 on HealthBench Professional, rank 15 of 31 models on the board. The last release of the Opus 4 series, superseded by Claude Opus 5 in July 2026. 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

rank	15 of 31
score	0.574
lab	Anthropic
context window	1.0M tokens
API price per 1M tokens	$5.00 in / $25.00 out
license	proprietary
source	System Card: Claude Sonnet 5 (system card)
released	2026-05-28

Position in the field

The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.129. Directly above sits GPT-5.6 Terra at 0.577. Directly below sits Grok 4.7 at 0.567. The scores on this page are compiled from published documents rather than from one controlled run, so a small gap between 2 models can reflect a difference in grader version, reasoning effort, or deployment setting as well as a difference in capability.

Source of this score

Read from System Card: Claude Sonnet 5 (system card, Anthropic, 2026-06-30). Vendor-reported. Confidence: verified. Configuration: Anthropic June evaluation; length-adjusted; adaptive max effort; Opus 4.8 grader; five trials; no tools or custom system prompt.

HealthBench Professional | Length-adjusted score (%) | Claude Opus 4.8 | 57.4%

p. 139, Figure 8.12.2.A, Claude Opus 4.8 bar (visually read printed labels). · full entry on the sources page

What does Claude Opus 4.8 (Opus 4.8 grader) score on HealthBench Professional?

Claude Opus 4.8 (Opus 4.8 grader) scores 0.574 on HealthBench Professional, which places it at rank 15 of 31 models on the board as of September 30, 2026. The number was read from System Card: Claude Sonnet 5, listed on the sources page.

How much does Claude Opus 4.8 (Opus 4.8 grader) cost to run?

Claude Opus 4.8 (Opus 4.8 grader) is priced at $5.00 per million input tokens and $25.00 per million output tokens through Anthropic's API.

Head to head

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

Claude Opus 4.8 (Opus 4.8 grader) vs GPT-6 Astra (Anthropic run)
0.574 vs 0.703 · GPT-6 Astra (Anthropic run) by 0.129
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Sonnet 5.5
0.574 vs 0.692 · Claude Sonnet 5.5 by 0.118
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Fable 5
0.574 vs 0.660 · Claude Fable 5 by 0.086
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Opus 5.5
0.574 vs 0.656 · Claude Opus 5.5 by 0.082
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-6 Astra
0.574 vs 0.647 · GPT-6 Astra by 0.073
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Fable 5 (September card)
0.574 vs 0.633 · Claude Fable 5 (September card) by 0.059
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Fable 5.1
0.574 vs 0.621 · Claude Fable 5.1 by 0.047
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-6 Luna
0.574 vs 0.608 · GPT-6 Luna by 0.034
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-6 Sol
0.574 vs 0.608 · GPT-6 Sol by 0.034
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.6 Sol
0.574 vs 0.605 · GPT-5.6 Sol by 0.031
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Opus 5
0.574 vs 0.598 · Claude Opus 5 by 0.024
Claude Opus 4.8 (Opus 4.8 grader) vs Muse Spark 1.1
0.574 vs 0.593 · Muse Spark 1.1 by 0.019
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Sonnet 5
0.574 vs 0.578 · Claude Sonnet 5 by 0.004
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.6 Terra
0.574 vs 0.577 · GPT-5.6 Terra by 0.003
Claude Opus 4.8 (Opus 4.8 grader) vs Grok 4.7
0.574 vs 0.567 · Claude Opus 4.8 (Opus 4.8 grader) by 0.007
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Opus 4.8
0.574 vs 0.558 · Claude Opus 4.8 (Opus 4.8 grader) by 0.016
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.6 Luna
0.574 vs 0.557 · Claude Opus 4.8 (Opus 4.8 grader) by 0.017
Claude Opus 4.8 (Opus 4.8 grader) vs Muse Spark
0.574 vs 0.541 · Claude Opus 4.8 (Opus 4.8 grader) by 0.033
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.6 Sol (August)
0.574 vs 0.540 · Claude Opus 4.8 (Opus 4.8 grader) by 0.034
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Opus 4.7
0.574 vs 0.519 · Claude Opus 4.8 (Opus 4.8 grader) by 0.055
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.5
0.574 vs 0.518 · Claude Opus 4.8 (Opus 4.8 grader) by 0.056
Claude Opus 4.8 (Opus 4.8 grader) vs Grok 4.6
0.574 vs 0.485 · Claude Opus 4.8 (Opus 4.8 grader) by 0.089
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.4
0.574 vs 0.481 · Claude Opus 4.8 (Opus 4.8 grader) by 0.093
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5
0.574 vs 0.462 · Claude Opus 4.8 (Opus 4.8 grader) by 0.112
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.2
0.574 vs 0.459 · Claude Opus 4.8 (Opus 4.8 grader) by 0.115
Claude Opus 4.8 (Opus 4.8 grader) vs Claude Sonnet 4.6
0.574 vs 0.442 · Claude Opus 4.8 (Opus 4.8 grader) by 0.132
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.6 Luna (August)
0.574 vs 0.441 · Claude Opus 4.8 (Opus 4.8 grader) by 0.133
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.1
0.574 vs 0.396 · Claude Opus 4.8 (Opus 4.8 grader) by 0.178
Claude Opus 4.8 (Opus 4.8 grader) vs GPT-5.5 Instant
0.574 vs 0.384 · Claude Opus 4.8 (Opus 4.8 grader) by 0.190
Claude Opus 4.8 (Opus 4.8 grader) vs MAI-Thinking-1
0.574 vs 0.350 · Claude Opus 4.8 (Opus 4.8 grader) by 0.224

Where the scores come from and how they are read is on the methodology page. The full ranking is on the leaderboard.