Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Knowledge benchmark
HealthBench length-adjusted score. Every model the catalog carries a published HealthBench (length-adjusted) value for, ranked by that value.
HealthBench score after applying a verbosity penalty to model responses. 8 models in the ModelScale catalog have a published value.
Each point is a model with a published HealthBench (length-adjusted) value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
8 models with a published HealthBench (length-adjusted) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Length-adjusted rubric score |
|---|---|---|---|
| 1 | Anthropic | 65.4 | |
| 2 | Anthropic | 60.6 | |
| 3 | OpenAI | 58.5 |
What the benchmark measures and how its value is published, from the source’s own record.
HealthBench (length-adjusted) (HealthBench length-adjusted score) is a knowledge benchmark published by HealthBench: Evaluating Large Language Models Towards Improved Human Health. HealthBench score after applying a verbosity penalty to model responses. Its evaluation set is described by the source as 5,000 multi-turn patient conversations. The source rates its difficulty as realistic healthcare conversations.
Each model's result is published by HealthBench: Evaluating Large Language Models Towards Improved Human Health as length-adjusted rubric score. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 8 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.
| 4 | OpenAI | 58.3 |
| 5 | Anthropic | 57.8 |
| 6 | OpenAI | 55.1 |
| 7 | OpenAI | 54.5 |
| 8 | OpenAI | 53.2 |
Evidence key: Observed
Rows are ordered by the value HealthBench: Evaluating Large Language Models Towards Improved Human Health published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.