Llama 3.1

Llama 3.1

Llama 3.1 is Meta’s open-weight family of language models you can run locally via Ollama or Hugging Face. Available in 8B, 70B and 405B sizes with up to 128K context.

DeveloperMeta
FamilyLlama
Parameters8B / 70B / 405B
Context window128K
LicenseLlama 3.1 Community License
Opennessopen-weight
VRAM / RAM8B ≈ 6–8GB VRAM (Q4) · 70B ≈ 40GB+
Ollamallama3.1

Capabilities

  • Generation & chat
  • Reasoning & coding
  • Multilingual

Sources