Compare AI models

Specifications
Parameters
16B314B
Context window
32k8k
Benchmarks
HumanEval
57.3%63.2%
MATH
27.9%23.9%
MMLU
55.7%73%
GSM8K
72%62.9%
Overview
CompanyDeepSeekSpaceXSI
Release dateMay 16 2024Nov 3 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V2 Lite and Grok‑1 are evenly matched across the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). Grok‑1 shipped 195 days before DeepSeek V2 Lite, so benchmark comparisons should account for the intervening progress.

DeepSeek V2 Lite has 16B parameters, while Grok‑1 has 314B. Context windows are 32k (DeepSeek V2 Lite) vs 8k (Grok‑1).

On HumanEval, Grok‑1 leads at 63.2% vs DeepSeek V2 Lite at 57.3%. On MATH, DeepSeek V2 Lite leads at 27.9% vs Grok‑1 at 23.9%. On MMLU, Grok‑1 leads at 73% vs DeepSeek V2 Lite at 55.7%. On GSM8K, DeepSeek V2 Lite leads at 72% vs Grok‑1 at 62.9%.