LLaMA at a glance
LLaMA is an AI apis & infrastructure tool. A foundational, 65-billion-parameter large language model by Meta.…
- Best for:
- Released July 2023 by Meta AI
- Pricing:
- Open Source
- Standout feature:
- Text generation and completion in 7B, 13B, and 70B parameter variants
- Main limitation:
- Context window limited to 4,096 tokens — well below GPT-4 or Claude contemporaries
About LLaMA
Llama 2 is Meta AI's open-weight large language model family released in July 2023, available in 7B, 13B, and 70B parameter sizes. It includes both pretrained base models and instruction-tuned chat variants (Llama-2-chat) trained with reinforcement learning from human feedback (RLHF). Unlike closed models, Llama 2 weights are freely downloadable, enabling self-hosted deployment on local hardware or private cloud infrastructure. It targets researchers, developers, and enterprises building AI applications who need full control over their stack, data privacy, or cost efficiency. The 70B chat model is competitive with many commercial models on standard benchmarks.
Key Features of LLaMA
Here's what LLaMA actually does — the capabilities our editorial team and AI research surfaced from the product, documentation, and user reports:
