I kept running into new benchmarks while doing benchmark research, so I built a crawler that continuously collects benchmark-related signals from across the web. It pulls evidence from arXiv, GitHub, Hugging Face, OpenAlex, OpenReview, first-party lab feeds, Brave Search, Semantic Scholar, Hacker News, and more every day, and keeps updating.
简体中文版本请见 README.zh-CN.md.
- Open the dashboard — today's insights, trends, popular benchmarks, model-card adoption, and more
- Subscribe via RSS — get new benchmark signals every day
- Download all curated data — export the full machine-readable corpus in one click
- Contribute — add benchmarks, model cards, sources, or fixes
If this is useful, star the repo.
- Scoring rubric:
src/benchmark_radar/rubric.py - Model-card adoption data:
data/model_cards.yml - Public corpus schema:
docs/cumulative-corpus.schema.json - Configuration:
config.yml - Run locally:
python -m pip install -e '.[dev]' && benchmark-radar - Support / bugs: open an issue
- Contact: @ktwu01
- License: MIT