Claude Sonnet 5 on HealthBench Professional
Claude Sonnet 5 scores 0.578 on HealthBench Professional, rank 19 of 43 models on the board. Anthropic's mid-tier model, positioned for speed at near-Opus quality. 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.
Where it sits on the board
The models ranked around Claude Sonnet 5, by length-adjusted score. Differences are against Claude Sonnet 5.
| Rank | Model | Length-adjusted | Before the adjustment | Difference |
|---|---|---|---|---|
| 16 | 0.605 | – | +0.027 | |
| 17 | 0.598 | 0.734 | +0.020 | |
| 18 | 0.593 | – | +0.015 | |
| 19 | 0.578 | 0.624 | – | |
| 20 | 0.577 | 0.624 | -0.001 | |
| 21 | 0.574 | – | -0.004 | |
| 22 | 0.567 | – | -0.011 |
Result and API facts
| Rank | 19 of 43 |
|---|---|
| Length-adjusted score | 0.578 |
| Score before the length adjustment | 0.624 |
| Lab | Anthropic |
| Context window | 1.0M tokens |
| API price per 1M tokens | $2.00 in / $10.00 out |
| License | proprietary |
| Source | System Card: Claude Sonnet 5 (system card) |
| Released | 2026-06-30 |
Source of this score
Read from System Card: Claude Sonnet 5 (system card, Anthropic, 2026-06-30). Vendor-reported. Confidence: verified. Configuration: length-adjusted, Anthropic protocol: adaptive thinking at max effort, Claude Opus 4.8 grader, averaged over 5 trials, no tools or custom system prompt (raw 62.4%)..
HealthBench Professional 57.8 44.2 51.8 -
Position in the field
The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.125. Directly above sits Muse Spark 1.1 at 0.593. Directly below sits GPT-5.6 Terra at 0.577. 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.
What does Claude Sonnet 5 score on HealthBench Professional?
Claude Sonnet 5 scores 0.578 on HealthBench Professional, which places it at rank 19 of 43 models on the board as of November 2, 2026. The number was read from System Card: Claude Sonnet 5, listed on the sources page.
How much does Claude Sonnet 5 cost to run?
Claude Sonnet 5 is priced at $2.00 per million input tokens and $10.00 per million output tokens through Anthropic's API.
Head to head
Score-difference matrixPairings with a dedicated comparison page are linked; every other difference is in the score-difference matrix.
- Claude Sonnet 5 vs GPT-6 Astra (Anthropic run)
- Claude Sonnet 5 vs Claude Sonnet 5.5
- Claude Sonnet 5 vs Claude Fable 5
- Claude Sonnet 5 vs Claude Opus 5.5
- Claude Sonnet 5 vs GPT-6 Luna (Anthropic run)
- Claude Sonnet 5 vs Claude Haiku 5.5
- Claude Sonnet 5 vs GPT-6 Astra
- Claude Sonnet 5 vs GPT-6.1 Sol
- Claude Sonnet 5 vs Claude Fable 5 (September card)
- Claude Sonnet 5 vs GPT-6.1 Terra
- Claude Sonnet 5 vs Muse Wonder Full
- Claude Sonnet 5 vs Claude Fable 5.1
- Claude Sonnet 5 vs GPT-6 Luna
- Claude Sonnet 5 vs GPT-6 Sol
- Claude Sonnet 5 vs GPT-5.6 Sol
- Claude Sonnet 5 vs Muse Wonder
- Claude Sonnet 5 vs Claude Opus 5
- Claude Sonnet 5 vs Muse Spark 1.1
- Claude Sonnet 5 vs Claude Opus 4.8 (Opus 4.8 grader)
- Claude Sonnet 5 vs Grok 4.7
- Claude Sonnet 5 vs Claude Opus 4.8
- Claude Sonnet 5 vs GPT-5.6 Luna
- Claude Sonnet 5 vs Gemini 4 Argon
- Claude Sonnet 5 vs Muse Spark
- Claude Sonnet 5 vs GPT-5.6 Sol (August)
- Claude Sonnet 5 vs Kimi K3
- Claude Sonnet 5 vs GLM 5.3
- Claude Sonnet 5 vs Claude Opus 4.7
- Claude Sonnet 5 vs GPT-5.5
- Claude Sonnet 5 vs Gemini 3.8 Flash
- Claude Sonnet 5 vs DeepSeek V4.1 Flash
- Claude Sonnet 5 vs Grok 4.6
- Claude Sonnet 5 vs GPT-5.4
- Claude Sonnet 5 vs GPT-5
- Claude Sonnet 5 vs GPT-5.2
- Claude Sonnet 5 vs Claude Sonnet 4.6
- Claude Sonnet 5 vs GPT-5.6 Luna (August)
- Claude Sonnet 5 vs GPT-5.1
- Claude Sonnet 5 vs GPT-5.5 Instant
- Claude Sonnet 5 vs MAI-Thinking-1
- Claude Sonnet 5 vs Claude Haiku 4.5
Where the scores come from and how they are read is on the methodology page. The full ranking is on the leaderboard.