Sources / Press and newsletters

LAION-5B: An open large-scale dataset for training next generation image-text models

1 record was read from this source, dated 31 Mar 2022.

Last checked 11 Oct 2026

What was read from it

Records read from this source
Record	Kind	As stated	Dated
LAION-5B contains 5.85 billion CLIP-filtered image-text pairs: 2.3 billion in English and 2.2 billion in more than 100 other languages.	Supply fact	5,85B	31 Mar 2022