data-datasets
Hugging Face Datasets
The largest hub of machine-learning datasets (and models), each carrying rich metadata. A keyless API serves search results, which our site queries directly, and everything is browsable on the website. The default place to look when you need a published ML dataset or its documentation.
Why it’s useful & how it works
Hugging Face is the central hub for the machine-learning community to publish and discover datasets used to train, fine-tune, and evaluate AI models. Each dataset entry carries a data card describing its contents, size, languages, licenses, and intended use, making it easy to evaluate whether a dataset suits your needs before downloading. Echo queries the keyless Hugging Face API so you can find datasets by topic, modality, or task; you then explore and download on huggingface.co.
What’s inside
More than 730,000 datasets as of mid-2026, covering text, images, audio, video, code, and multimodal formats across dozens of languages. The hub also hosts over 2 million model checkpoints linked to these datasets.
API access
https://huggingface.co/api/datasets?search= (keyless read)
What we measured
Our own probes, not the archive’s own claims. Re-run periodically; every reading below is dated.
Reachability
- Direct request
- Responded HTTP 200 822 ms
- Through a datacenter proxy
- Responded HTTP 200 1.1 s
- API, direct
- Responded HTTP 200 199 ms
- API, through a proxy
- Responded HTTP 200 713 ms
Did a real search return anything?
Inconclusive · median 11.5 s · down during the run
Slow: budget for long waits, or call it asynchronously.
Our proxy pool could not connect, so this query never reached the archive: inconclusive, not a failure of the archive.
Benchmark coverage
Found 9 of 10 test items (90%) in the Datasets cluster. Median response 757 ms.
By how well known the item is: 3/3 well-known, 6/6 mid, 0/1 obscure .
One item it found: ILSVRC/imagenet-1k
Reachability measured 2026-08-22. Search checked 2026-08-22. Benchmark run 2026-08-22.
Access
Programmatic API access (a key may be required, see the API tag).