Falcon 2: An 11B parameter pretrained language model and VLM, trained on over 5000B tokens and 11 languages
11B parameters. 5 trillion tokens. 11 languages. Falcon 2 just reset the efficiency bar for open-weight models.

Why it matters
Falcon 2 challenges the scale-first narrative by delivering competitive performance at 11B parameters across multilingual and vision tasks, directly threatening the cost assumptions of larger proprietary models.
The key facts
6 to know11B parameter architecture
5000B+ tokens training data
11-language multilingual support
Vision language model (VLM) capability
Open-weight release on Hugging Face
May 2024 release date
Go to the source
Hugging Face Bloghuggingface.co