What it measures
- Long-context applications
- Retrieval and reasoning
- Task-level robustness
HELMET
A long-context benchmark suite focused on practical application-style tasks.
How to read the score
score; higher is better.
Sourced Ranking
| Rank | Model | Provider | Score | Source | Source Type | Verified |
|---|
Limitations
- Track subtask mix and spreadsheet version.
- The arXiv paper is static; the project page and spreadsheet may update.
- Official page includes overall results and spreadsheet links.
leaderboard
Princeton HELMET project page and spreadsheet
The arXiv paper is static; the project page and spreadsheet may update. Official page includes overall results and spreadsheet links.
paper
HELMET paper or specification
Use for benchmark definition, not necessarily for latest model scores.
third-party
HELMET dataset or code
Use for dataset, code, or implementation details; score freshness depends on the benchmark.