HealthBench Professional

OpenAI logoGPT-6 Sol on HealthBench Professional

Rank 14 of 43, updated November 2, 2026.

GPT-6 Sol scores 0.608 on HealthBench Professional, rank 14 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.

14
rank of 43
0.608
length-adjusted score
0.595
score before the length adjustment
0.095
gap to the leader, GPT-6 Astra (Anthropic run)
1.05M
context window, tokens
$2.00 / $10.00
input / output price per 1M tokens

Where it sits on the board

The models ranked around GPT-6 Sol, by length-adjusted score. Differences are against GPT-6 Sol.

RankModelLength-adjustedBefore the adjustmentDifference
11Meta logoMuse Wonder Full0.623–+0.015
12Anthropic logoClaude Fable 5.10.6210.742+0.013
13OpenAI logoGPT-6 Luna0.6080.6120.000
14OpenAI logoGPT-6 Sol0.6080.595–
15OpenAI logoGPT-5.6 Sol0.6050.641-0.003
16Meta logoMuse Wonder0.605–-0.003
17Anthropic logoClaude Opus 50.5980.734-0.010

Result and API facts

Rank14 of 43
Length-adjusted score0.608
Score before the length adjustment0.595
LabOpenAI
Context window1.05M tokens
API price per 1M tokens$2.00 in / $10.00 out
Licenseproprietary
SourceGPT-6 Astra System Card - HealthBench (Deployment Safety Hub) (system card)
Released2026-09-22

Source of this score

Read from GPT-6 Astra System Card - HealthBench (Deployment Safety Hub) (system card, OpenAI, 2026-09-03). Vendor-reported. Confidence: verified. Configuration: OpenAI; length-adjusted; maximum reasoning effort; 60.8 (59.5, 1573) (adjusted, raw, mean response characters).

Section 11.4.1, Table 29, HealthBench Professional length-adjusted, GPT-6 Sol column; September 22 correction.full entry on the sources page
Evaluation | GPT-5.5 | GPT-5.6 Sol | GPT-5.6 Terra | GPT-5.6 Luna | GPT-6 Astra | GPT-6 Sol | GPT-6 Luna HealthBench Professional length-adjusted | 51.8 (57.2, 3818) | 60.5 (64.1, 3228) | 57.7 (62.4, 3618) | 55.7 (59.8, 3389) | 64.7 (68.2, 3185) | 60.8 (59.5, 1573) | 60.8 (61.2, 2119)

Position in the field

The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.095. Directly above sits GPT-6 Luna at 0.608. Directly below sits GPT-5.6 Sol at 0.605. 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 GPT-6 Sol score on HealthBench Professional?

GPT-6 Sol scores 0.608 on HealthBench Professional, which places it at rank 14 of 43 models on the board as of November 2, 2026. The number was read from GPT-6 Astra System Card - HealthBench (Deployment Safety Hub), listed on the sources page.

How much does GPT-6 Sol cost to run?

GPT-6 Sol is priced at $2.00 per million input tokens and $10.00 per million output tokens through OpenAI's API.

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

  • GPT-6 Sol vs GPT-6 Astra (Anthropic run)
    0.608 vs 0.703, GPT-6 Astra (Anthropic run) by 0.095
  • GPT-6 Sol vs Claude Sonnet 5.5
    0.608 vs 0.692, Claude Sonnet 5.5 by 0.084
  • GPT-6 Sol vs Claude Fable 5
    0.608 vs 0.660, Claude Fable 5 by 0.052
  • GPT-6 Sol vs Claude Opus 5.5
    0.608 vs 0.656, Claude Opus 5.5 by 0.048
  • GPT-6 Sol vs GPT-6 Luna (Anthropic run)
    0.608 vs 0.655, GPT-6 Luna (Anthropic run) by 0.047
  • GPT-6 Sol vs Claude Haiku 5.5
    0.608 vs 0.648, Claude Haiku 5.5 by 0.040
  • GPT-6 Sol vs GPT-6 Astra
    0.608 vs 0.647, GPT-6 Astra by 0.039
  • GPT-6 Sol vs GPT-6.1 Sol
    0.608 vs 0.642, GPT-6.1 Sol by 0.034
  • GPT-6 Sol vs Claude Fable 5 (September card)
    0.608 vs 0.633, Claude Fable 5 (September card) by 0.025
  • GPT-6 Sol vs GPT-6.1 Terra
    0.608 vs 0.627, GPT-6.1 Terra by 0.019
  • GPT-6 Sol vs Muse Wonder Full
    0.608 vs 0.623, Muse Wonder Full by 0.015
  • GPT-6 Sol vs Claude Fable 5.1
    0.608 vs 0.621, Claude Fable 5.1 by 0.013
  • GPT-6 Sol vs GPT-6 Luna
    0.608 vs 0.608, GPT-6 Luna by 0.000
  • GPT-6 Sol vs GPT-5.6 Sol
    0.608 vs 0.605, GPT-6 Sol by 0.003
  • GPT-6 Sol vs Muse Wonder
    0.608 vs 0.605, GPT-6 Sol by 0.003
  • GPT-6 Sol vs Claude Opus 5
    0.608 vs 0.598, GPT-6 Sol by 0.010
  • GPT-6 Sol vs Muse Spark 1.1
    0.608 vs 0.593, GPT-6 Sol by 0.015
  • GPT-6 Sol vs Claude Sonnet 5
    0.608 vs 0.578, GPT-6 Sol by 0.030
  • GPT-6 Sol vs GPT-5.6 Terra
    0.608 vs 0.577, GPT-6 Sol by 0.031
  • GPT-6 Sol vs Claude Opus 4.8 (Opus 4.8 grader)
    0.608 vs 0.574, GPT-6 Sol by 0.034
  • GPT-6 Sol vs Grok 4.7
    0.608 vs 0.567, GPT-6 Sol by 0.041
  • GPT-6 Sol vs Claude Opus 4.8
    0.608 vs 0.558, GPT-6 Sol by 0.050
  • GPT-6 Sol vs GPT-5.6 Luna
    0.608 vs 0.557, GPT-6 Sol by 0.051
  • GPT-6 Sol vs Gemini 4 Argon
    0.608 vs 0.551, GPT-6 Sol by 0.057
  • GPT-6 Sol vs Muse Spark
    0.608 vs 0.541, GPT-6 Sol by 0.067
  • GPT-6 Sol vs GPT-5.6 Sol (August)
    0.608 vs 0.540, GPT-6 Sol by 0.068
  • GPT-6 Sol vs Kimi K3
    0.608 vs 0.532, GPT-6 Sol by 0.076
  • GPT-6 Sol vs GLM 5.3
    0.608 vs 0.528, GPT-6 Sol by 0.080
  • GPT-6 Sol vs Claude Opus 4.7
    0.608 vs 0.519, GPT-6 Sol by 0.089
  • GPT-6 Sol vs GPT-5.5
    0.608 vs 0.518, GPT-6 Sol by 0.090
  • GPT-6 Sol vs Gemini 3.8 Flash
    0.608 vs 0.517, GPT-6 Sol by 0.091
  • GPT-6 Sol vs DeepSeek V4.1 Flash
    0.608 vs 0.507, GPT-6 Sol by 0.101
  • GPT-6 Sol vs Grok 4.6
    0.608 vs 0.485, GPT-6 Sol by 0.123
  • GPT-6 Sol vs GPT-5.4
    0.608 vs 0.481, GPT-6 Sol by 0.127
  • GPT-6 Sol vs GPT-5
    0.608 vs 0.462, GPT-6 Sol by 0.146
  • GPT-6 Sol vs GPT-5.2
    0.608 vs 0.459, GPT-6 Sol by 0.149
  • GPT-6 Sol vs Claude Sonnet 4.6
    0.608 vs 0.442, GPT-6 Sol by 0.166
  • GPT-6 Sol vs GPT-5.6 Luna (August)
    0.608 vs 0.441, GPT-6 Sol by 0.167
  • GPT-6 Sol vs GPT-5.1
    0.608 vs 0.396, GPT-6 Sol by 0.212
  • GPT-6 Sol vs GPT-5.5 Instant
    0.608 vs 0.384, GPT-6 Sol by 0.224
  • GPT-6 Sol vs MAI-Thinking-1
    0.608 vs 0.350, GPT-6 Sol by 0.258
  • GPT-6 Sol vs Claude Haiku 4.5
    0.608 vs 0.322, GPT-6 Sol by 0.286

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