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
Every model the catalog carries a published HLE-Verified value for, ranked by that value.
A verified and revised Humanity's Last Exam set that removes uncertain items and repairs fixable questions before evaluating frontier expert reasoning. 6 models in the ModelScale catalog have a published value.
Each point is a model with a published HLE-Verified value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
6 models with a published HLE-Verified 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 | Full-set accuracy |
|---|---|---|---|
| 1 | 54.9 | ||
| 2 | OpenAI | 54.5 | |
| 3 | Anthropic | 54.4 |
What the benchmark measures and how its value is published, from the source’s own record.
HLE-Verified is a knowledge benchmark published by HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam. A verified and revised Humanity's Last Exam set that removes uncertain items and repairs fixable questions before evaluating frontier expert reasoning. Its evaluation set is described by the source as 1,811 verified or revised expert questions. The source rates its difficulty as frontier multidisciplinary expert reasoning.
Each model's result is published by HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam as full-set accuracy. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 6 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 | 53.6 |
| 5 | OpenAI | 51.1 |
| 6 | Anthropic | 31.0 |
Evidence key: Observed
Rows are ordered by the value HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam 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.