HealthBench Professional

Meta logoMuse Wonder on HealthBench Professional

Rank 16 of 43, updated November 2, 2026.

Muse Wonder scores 0.605 on HealthBench Professional, rank 16 of 43 models on the board. 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.

16
rank of 43
0.605
length-adjusted score
–
score before the length adjustment, not printed in the source
0.098
gap to the leader, GPT-6 Astra (Anthropic run)
1M
context window, tokens
–
no published list pricing

Where it sits on the board

The models ranked around Muse Wonder, by length-adjusted score. Differences are against Muse Wonder.

RankModelLength-adjustedBefore the adjustmentDifference
13OpenAI logoGPT-6 Luna0.6080.612+0.003
14OpenAI logoGPT-6 Sol0.6080.595+0.003
15OpenAI logoGPT-5.6 Sol0.6050.6410.000
16Meta logoMuse Wonder0.605––
17Anthropic logoClaude Opus 50.5980.734-0.007
18Meta logoMuse Spark 1.10.593–-0.012
19Anthropic logoClaude Sonnet 50.5780.624-0.027

Result and API facts

Rank16 of 43
Length-adjusted score0.605
Score before the length adjustmentnot printed in the source
LabMeta
Context window1M tokens
API price per 1M tokensno published list pricing
Licenseproprietary
SourceMuse Wonder Model Card (model card)
Released2026-10-12

Source of this score

Read from Muse Wonder Model Card (model card, Meta, 2026-10-12). Vendor-reported. Confidence: verified. Configuration: Meta internal run; length-normalized weighted rubric score, GPT-5.4 grader, no tools, max reasoning effort (Muse Wonder Model Card p. 3 table, Muse Wonder (max) column).

p. 3, results table (image), HealthBench Pro row, Muse Wonder (max) column; methodology p. 4full entry on the sources page
HealthBench Pro (Open-ended health queries) | Muse Wonder Full (max) 62.3 | Muse Wonder (max) 60.5 | Claude Fable 5.1 (max) 62.1 | GPT-6 Astra (max) 65.3 | Claude Opus 5.5 (max) 65.6 Methodology (p. 4): HealthBench Pro evaluates responses to 525 professional-grade health questions. Models answer without tools, and a GPT-5.4 grader scores each response against physician-written rubrics. We run the official tasks and grading in a common internal framework. The primary metric is the length-normalized weighted rubric score.

Position in the field

The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.098. Directly above sits GPT-5.6 Sol at 0.605. Directly below sits Claude Opus 5 at 0.598. 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 Muse Wonder score on HealthBench Professional?

Muse Wonder scores 0.605 on HealthBench Professional, which places it at rank 16 of 43 models on the board as of November 2, 2026. The number was read from Muse Wonder Model Card, listed on the sources page.

How much does Muse Wonder cost to run?

Meta has not published API pricing for Muse Wonder.

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

  • Muse Wonder vs GPT-6 Astra (Anthropic run)
    0.605 vs 0.703, GPT-6 Astra (Anthropic run) by 0.098
  • Muse Wonder vs Claude Sonnet 5.5
    0.605 vs 0.692, Claude Sonnet 5.5 by 0.087
  • Muse Wonder vs Claude Fable 5
    0.605 vs 0.660, Claude Fable 5 by 0.055
  • Muse Wonder vs Claude Opus 5.5
    0.605 vs 0.656, Claude Opus 5.5 by 0.051
  • Muse Wonder vs GPT-6 Luna (Anthropic run)
    0.605 vs 0.655, GPT-6 Luna (Anthropic run) by 0.050
  • Muse Wonder vs Claude Haiku 5.5
    0.605 vs 0.648, Claude Haiku 5.5 by 0.043
  • Muse Wonder vs GPT-6 Astra
    0.605 vs 0.647, GPT-6 Astra by 0.042
  • Muse Wonder vs GPT-6.1 Sol
    0.605 vs 0.642, GPT-6.1 Sol by 0.037
  • Muse Wonder vs Claude Fable 5 (September card)
    0.605 vs 0.633, Claude Fable 5 (September card) by 0.028
  • Muse Wonder vs GPT-6.1 Terra
    0.605 vs 0.627, GPT-6.1 Terra by 0.022
  • Muse Wonder vs Muse Wonder Full
    0.605 vs 0.623, Muse Wonder Full by 0.018
  • Muse Wonder vs Claude Fable 5.1
    0.605 vs 0.621, Claude Fable 5.1 by 0.016
  • Muse Wonder vs GPT-6 Luna
    0.605 vs 0.608, GPT-6 Luna by 0.003
  • Muse Wonder vs GPT-6 Sol
    0.605 vs 0.608, GPT-6 Sol by 0.003
  • Muse Wonder vs GPT-5.6 Sol
    0.605 vs 0.605, GPT-5.6 Sol by 0.000
  • Muse Wonder vs Claude Opus 5
    0.605 vs 0.598, Muse Wonder by 0.007
  • Muse Wonder vs Muse Spark 1.1
    0.605 vs 0.593, Muse Wonder by 0.012
  • Muse Wonder vs Claude Sonnet 5
    0.605 vs 0.578, Muse Wonder by 0.027
  • Muse Wonder vs GPT-5.6 Terra
    0.605 vs 0.577, Muse Wonder by 0.028
  • Muse Wonder vs Claude Opus 4.8 (Opus 4.8 grader)
    0.605 vs 0.574, Muse Wonder by 0.031
  • Muse Wonder vs Grok 4.7
    0.605 vs 0.567, Muse Wonder by 0.038
  • Muse Wonder vs Claude Opus 4.8
    0.605 vs 0.558, Muse Wonder by 0.047
  • Muse Wonder vs GPT-5.6 Luna
    0.605 vs 0.557, Muse Wonder by 0.048
  • Muse Wonder vs Gemini 4 Argon
    0.605 vs 0.551, Muse Wonder by 0.054
  • Muse Wonder vs Muse Spark
    0.605 vs 0.541, Muse Wonder by 0.064
  • Muse Wonder vs GPT-5.6 Sol (August)
    0.605 vs 0.540, Muse Wonder by 0.065
  • Muse Wonder vs Kimi K3
    0.605 vs 0.532, Muse Wonder by 0.073
  • Muse Wonder vs GLM 5.3
    0.605 vs 0.528, Muse Wonder by 0.077
  • Muse Wonder vs Claude Opus 4.7
    0.605 vs 0.519, Muse Wonder by 0.086
  • Muse Wonder vs GPT-5.5
    0.605 vs 0.518, Muse Wonder by 0.087
  • Muse Wonder vs Gemini 3.8 Flash
    0.605 vs 0.517, Muse Wonder by 0.088
  • Muse Wonder vs DeepSeek V4.1 Flash
    0.605 vs 0.507, Muse Wonder by 0.098
  • Muse Wonder vs Grok 4.6
    0.605 vs 0.485, Muse Wonder by 0.120
  • Muse Wonder vs GPT-5.4
    0.605 vs 0.481, Muse Wonder by 0.124
  • Muse Wonder vs GPT-5
    0.605 vs 0.462, Muse Wonder by 0.143
  • Muse Wonder vs GPT-5.2
    0.605 vs 0.459, Muse Wonder by 0.146
  • Muse Wonder vs Claude Sonnet 4.6
    0.605 vs 0.442, Muse Wonder by 0.163
  • Muse Wonder vs GPT-5.6 Luna (August)
    0.605 vs 0.441, Muse Wonder by 0.164
  • Muse Wonder vs GPT-5.1
    0.605 vs 0.396, Muse Wonder by 0.209
  • Muse Wonder vs GPT-5.5 Instant
    0.605 vs 0.384, Muse Wonder by 0.221
  • Muse Wonder vs MAI-Thinking-1
    0.605 vs 0.350, Muse Wonder by 0.255
  • Muse Wonder vs Claude Haiku 4.5
    0.605 vs 0.322, Muse Wonder by 0.283

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