Compare AI models

Specifications
Parameters
16B9B/27B
Context window
32k8k
API pricing
Cheapest input
—$0.65NextBit
Cheapest output
—$0.65NextBit
—$163NextBit · int4
Benchmarks
HumanEval
57.3%51.8%
MATH
27.9%42.3%
MMLU
55.7%75.2%
GSM8K
72%74%
Overview
CompanyDeepSeekGoogle
Release dateMay 16 2024Jun 27 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemma 2 leads DeepSeek V2 Lite on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). DeepSeek V2 Lite shipped 42 days before Gemma 2, so benchmark comparisons should account for the intervening progress.

DeepSeek V2 Lite has 16B parameters, while Gemma 2 has 9B/27B. Context windows are 32k (DeepSeek V2 Lite) vs 8k (Gemma 2).

On HumanEval, DeepSeek V2 Lite leads at 57.3% vs Gemma 2 at 51.8%. On MATH, Gemma 2 leads at 42.3% vs DeepSeek V2 Lite at 27.9%. On MMLU, Gemma 2 leads at 75.2% vs DeepSeek V2 Lite at 55.7%. On GSM8K, Gemma 2 leads at 74% vs DeepSeek V2 Lite at 72%.