Compare AI models

Specifications
Parameters
671B9B/27B
Context window
128k8k
API pricing
Cheapest input
$0.256DeepInfra$0.65NextBit
Cheapest output
$0.712DeepInfra$0.65NextBit
$103StreamLake$163NextBit · int4
Benchmarks
MMLU
88.5%75.2%
Overview
CompanyDeepSeekGoogle
Release dateDec 26 2024Jun 27 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V3 leads Gemma 2 on 1 of the 1 benchmark they both report (MMLU). Gemma 2 shipped 182 days before DeepSeek V3, so benchmark comparisons should account for the intervening progress.

DeepSeek V3 has 671B parameters, while Gemma 2 has 9B/27B. Context windows are 128k (DeepSeek V3) vs 8k (Gemma 2).

On MMLU, DeepSeek V3 leads at 88.5% vs Gemma 2 at 75.2%.