Skip to content
AI Model Radar

Llama 3.1 8B Instruct

meta-llama/Llama-3.1-8B-Instruct
Open Source · chat · open-weights
GA Alert me on changes
Context
131.1K
Max output
16.4K
Input $/1M
$0.02
Output $/1M
$0.03
Modalities
text
Released
18 Jul 2024
License: llama3.1 · meta-llama/Llama-3.1-8B-Instruct
AI summary
● machine-written

Meta releases Llama 3.1 8B Instruct, open-weight chat model

Llama 3.1 8B Instruct is an 8-billion-parameter language model from Meta optimized for dialogue and instruction-following tasks. The model features a 131,072-token context window, supports tool use and multilingual capabilities, and is available through multiple providers at varying price points. Released in July 2024, it operates under the Llama 3.1 Community License and has demonstrated competitive performance compared to larger proprietary models.

What's new
  • Context window of 131,072 tokens enables processing of long documents
  • Maximum output generation of 16,384 tokens
  • Available via 9 providers including Lambda, Groq, Sambanova, and Cerebras
  • Knowledge cutoff: December 2023
  • Open-weights distribution allows self-hosting
Best for
Dialogue and conversational applicationsTool-calling and function integrationMultilingual text generationInstruction-following tasks
Sources

Source: https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct