Meta: Llama 3.3 70B Instruct

Made by meta-llama. Open weights

Input price
$0.10
per 1M tokens
Output price
$0.32
per 1M tokens
Context window
131K
tokens
Quality
1,274
#273 on LMArena
Size
71B
parameters
Downloads
354.9K
on Hugging Face, last 30 days

Can I run it locally?

Estimate for the Q4_K_M version with an 8K context. Pick your own card and settings in the calculator.

  • NVIDIA GeForce RTX 40608 GBRuns, but slowly1.0 to 1.4 tokens/s
  • NVIDIA GeForce RTX 306012 GBRuns, but slowly1.1 to 1.5 tokens/s
  • NVIDIA GeForce RTX 4060 Ti 16 GB16 GBRuns, but slowly1.2 to 1.6 tokens/s
  • NVIDIA GeForce RTX 409024 GBRuns, but slowly1.7 to 2.4 tokens/s
  • NVIDIA GeForce RTX 509032 GBRuns, but slowly2.8 to 3.9 tokens/s
  • Apple M1 Pro 32 GB32 GBWill not run—
Check my computer

Prices: OpenRouter, refreshed 8 Oct 2026. Quality: LMArena (CC BY 4.0), leaderboard of 2 Oct 2026. Model page on Hugging Face