Compare AI models

Specifications
Parameters
236B7B/13B/33B/65B
Context window
128k2k
Benchmarks
HumanEval
84.76%23.7%
MATH
71.02%10.6%
Overview
CompanyDeepSeekMeta
Release dateJun 28 2024Feb 24 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V2-0628 leads LLaMA 1 on 2 of the 2 benchmarks they both report (HumanEval, MATH). LLaMA 1 shipped 490 days before DeepSeek V2-0628, so benchmark comparisons should account for the intervening progress.

DeepSeek V2-0628 has 236B parameters, while LLaMA 1 has 7B/13B/33B/65B. Context windows are 128k (DeepSeek V2-0628) vs 2k (LLaMA 1).

On HumanEval, DeepSeek V2-0628 leads at 84.76% vs LLaMA 1 at 23.7%. On MATH, DeepSeek V2-0628 leads at 71.02% vs LLaMA 1 at 10.6%.