Llama 3.1 8B

Llama 3.1 8B is Meta AI's compact open-weights language model released in July 2024 as part of the Llama 3.1 family. Designed for efficiency, local deployment, and low-latency developer applications, Llama 3.1 8B expands the context window from 8k in previous releases to 131,072 tokens (128k), matching the long-context capabilities of its 70B and 405B sibling models. The model was pre-trained on over 15 trillion tokens of multilingual and code data. Meta officially benchmarked and instruction-tuned Llama 3.1 8B for 8 core languages (English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai). It supports function calling, tool use, retrieval-augmented generation (RAG), and zero-shot reasoning in resource-constrained environments. As an open-weights model governed by the Meta Llama 3.1 Community License, Llama 3.1 8B can be deployed locally using Ollama, vLLM, LM Studio, or ONNX runtimes. While Meta's official model cards define the 128k context window, max output limits are host- and inference-engine specific rather than hard-capped in model weights.