Llama
Meta's open-weight model family, from small on-device models to very large ones.
ChatLong contextMultilingual
Overview
Llama is Meta's family of open-weight large language models. It covers everything from small models that run on a single GPU to very large multilingual models, and many other open models are fine-tuned from it.
The Llama Community Licence allows commercial use, but adds conditions: very large services (over 700 million monthly users) need a separate licence from Meta, and use must follow Meta's acceptable use policy.
Models in this family
| Model | Released | Parameters | Context |
|---|---|---|---|
| Llama 3.3 70B Instruct | 6 Dec 2024 | 70B | 128K tokens |
| Llama 3.1 8B Instruct | 23 Jul 2024 | 8B | 128K tokens |