Meta

Llama 3.3 Instruct 70B

Dec 6, 2024Open weights

Compare

Intelligence

7.7

#520 of 744Rank 520 of 744 measured models

Coding

11.9

#261 of 298Rank 261 of 298 measured models

Output speed

92.7 t/s

#190 of 403Rank 190 of 403 measured models

Price/1M

$0.712

#331 of 659Rank 331 of 659 measured models

Context window

128K

#773 of 1199Rank 773 of 1199 measured models

Llama 3.3 Instruct 70B scores 7.7 on the Artificial Analysis Intelligence Index, ranking 520th of 744 measured models. The closest models on this index are GPT-4o (Aug '24) (7.7), Meta: Llama 3.3 70B Instruct (7.7), and Qwen2.5 72B Instruct (7.7). At $0.712 per million tokens (3:1 input/output blended price), it is cheaper than GPT-4o (Aug '24) ($4.38), more expensive than Meta: Llama 3.3 70B Instruct ($0.64) and Qwen2.5 72B Instruct ($0.37). With an output speed of 92.7 t/s, it is slower than GPT-4o (Aug '24) (147 t/s).