How to use Llama
Download the weights from Meta or Hugging Face and run them locally or through a cloud provider. Choose the lighter, long-context model for efficiency or the larger model for stronger quality. Review the Llama Community License before commercial use, especially at large scale.
Best for
Self-hosted and fine-tuned deployments where you control the weights and data. Very-long-context work. Cost control at scale.
How it behaves
Released under Meta's Llama Community License, free for research and most commercial use, with a separate agreement required for the largest platforms. It uses a mixture-of-experts design, and you host and prompt it yourself rather than paying a subscription.
Tips
Self-host when you need full control of data and weights. Pick the model size to fit your hardware budget. Check the license terms if you operate at very large scale.
Pricing
Free to download and self-host. You pay only for your own compute, with no subscription or license fee for the weights.