# Llama 3.1 8B requirements

Updated: 2026-06-15

llama family, 8B params, released 2024-07.

- Q4_K_M: 4.92 GB on disk; ~6.4 GB to run at 4k context
- Q8_0: 8.54 GB
- Context: 128k
- Run: `ollama run llama3.1:8b`
- Runs on 32 of 39 tracked devices

Sources: https://ollama.com/library/llama3.1, https://ollama.com/library/llama3.1/tags, https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF, https://lmarena.ai/leaderboard
More: https://localmodel.run/model/llama-3.1-8b